[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference heron::euro_swas_ai

Title:Europe-Swas-Artificial-Intelligence
Moderator:HERON::BUCHANAN
Created:Fri Jun 03 1988
Last Modified:Thu Aug 04 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:442
Total number of notes:1429

67.0. "FYI - US / DEC / AI TRAINING OFFERINGS" by HERON::ROACH (TANSTAAFL !) Tue Jan 24 1989 20:56

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     17-Jan-1989 07:12pm CET
                                        From:     HELLER
                                                  HELLER@AIADM@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: FYI

           INTELLIGENT SYSTEMS TECHNOLOGY TRAINING PROGRAMS
                          Q3/4 SCHEDULE
                            FY' 89

Listed below is the Q3/4 schedule for all the programs and courses offered 
by the ISTT group.  All courses are held at the AI Training Center on 
Donald Lynch Blvd in Marlboro(DLB12).  To register, please contact Louise 
King at AIADM::KING.  For more information, contact Janice Heller at
AIADM::HELLER.  Thanks very much.



KNOWLEDGE ENGINEERING CURRICULUM       8 1/2 weeks      starts Feb.20th
                                                        starts April 24th

INTRODUCTION TO AI IN BUSINESS         4 days           Feb. 20-23
                                                        April 24-27

   Overview of AI /Expert Systems      1 day            Feb 20th, 
                                                        April 24th
               
MANAGING EXPERT SYSTEMS                4 days           May 15-18



                       ADVANCED AI TECHNICAL COURSES

COURSE                   LENGTH      OFFERED           INSTRUCTOR

*Blackboard-based        3 days     Feb.27-March 1     Dr. Daniel Corkill
AI Systems                                             Dr. Victor Lesser

Advanced VAX OPS5        5 days     April 3-7          Tom Cooper
                                                       Lesley Chesler

Knowledge Craft          10 days    April 24-          Carnegie
                                    May 5                Group, Inc.

Advanced Knowledge       2.5 hrs/wk April 25,28,       Dr. Edwina
 Representation          for 6 wks  May 2,9,16,          Rissland
                         (1-3:30)   23,30

Knowledge                3 days     May 22-24          Dr. Raoul Smith
Representation
and Reasoning

Object Oriented          5 days     June 12-16         Dr. Stan Zdonik
Technology
and it's Role in
Database Systems

*Integrating             3 days     June 19-21         John Arnold
AI-Based
Systems


*New Course
T.RTitleUserPersonal
Name
DateLines
67.3Seeing is believingULYSSE::MOEDERTue Mar 07 1989 15:335
    Hi all,
    
    I was lucky enough to attend - and have copies of their presentations.
    If you want a copy, give a call!
    
67.4Integrating Knowledge-Based Systems CourseHERON::ROACHTANSTAAFL !Tue May 23 1989 13:0494
Printed by: Pat Roach                                    Document Number: 007267
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-May-1989 04:00pm ETE
                                        From:     DOUGLASS
                                                  DOUGLASS@AITG@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Integrating Knowledge-Based Systems Course

From:	AITG::KAEWERT      19-MAY-1989 15:52:40.83
To:	DOUGLASS
CC:	
Subj:	Did you receive this? Your famous mailing list might like to see this.

From:	ISTG::AIADM::KING "Louise King 291-8603  18-May-1989 1026" 18-MAY-1989 10:44:41.35
To:	@ALUMNI.DIS,@ADVANCED.DIS,KING        
CC:	
Subj:	Announcing a new advanced course -


********************************************************************************
	           INTEGRATING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

       DATE:	June 19th and 21st, 1989

   LOCATION:	DLB12-1 - Atlantic Conference Room
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:    The successful delivery of an application that uses artificial
                intelligence (AI) tools and techniques frequently depends
                on the ability of the system to coexist with other
                applications and existing data.  Using examples and
                exercises from case studies, this course presents
                guidelines and methods for integrating knowledge-based or
                expert systems into the application environment and easing
                a system's transition from prototype to production.  The
                course will highlight issues in project planning, system
                life cycle, knowledge representation, and implementation
                that must be addressed to successfully integrate a
                knowledge-based system into a business application
                portfolio.

      GOALS:    Upon completion of the course, participants will be able
                to:

                    - Define integration requirements for knowledge-based
                      applications

                    - Incorporate integration issues into the project plan

                    - Define the relationship (and a transformation)
                      between a relational data definition and a
                      knowledge representation

                    - Write simple VAX LISP programs that access non-LISP
                      routines and/or data files

                    - Identify trends toward the integration of databases
                      and AI applications
 
 PRESENTER:	John Arnold

  AUDIENCE:     Application developers, programmers, or software engineers
                involved with the delivery of knowledge-based or other AI
                systems. 
                Project leaders or technical managers of applications that
                have or are considering a knowledge-based component.


LENGTH & FORMAT:    3 days, Lecture/Lab


PREREQUISITES:  - Completion of the Expert Systems Training Program or
                  equivalent

                - Applications development experience

                - Familiarity with LISP programming

        COST:   $1,500 non manufacturing employees

    REGISTER:	Send mail to AIADM::KING.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.5SPECIAL PRESENTATION BY CAN BAKYAN ON - THURSDAY, MHERON::ROACHTANSTAAFL !Fri May 26 1989 17:5483
Printed by: Pat Roach                                    Document Number: 007352
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     26-May-1989 12:43pm ETE
                                        From:     GENTILE
                                                  GENTILE@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: SPECIAL PRESENTATION BY CAN BAKYAN ON - THURSDAY, MAY 25

                   SPECIAL PRESENTATION - THURSDAY, MAY 25                   
                                 2:00-3:00            
                            ATLANTIC CR (DLB12)         
 

 Can Bakyan from the CMU Center for Integrated Decision Systems (CIMDS)
will present a talk on the work he is doing in automated space planning.
This talk is sponsored by Chuck Marshall, ASTG/ISTG. Can (pronounced"John")
will be available for discussions after the talk - please notify Theresa
AIADM::Gentile if you wish separate time with Can.


 

		               ABSTRACT


Spatial Planning as a Constraint Satisfaction Problem.
 
WRIGHT is a CAD system for designing two dimensional layouts consisting 
of rectangles. This problem arises in space planning, i.e. the design of 
floorplans, arrangement of equipment in rooms, and site planning.
 
A layout is represented as a constraint satisfaction problem consisting of
variables and constraints between them. Search generates different CSP's by
adding constraints and local propagation of constraints within each CSP
eliminates inconsistent values.
 
The objects being configured are called design units. A design unit 
is a rectangle with an orientation. A rectangle is a structured variable 
made of two vertical lines, two horizontal lines, length, width and area.
A layout is represented as a CSP. The set of variables are the orientations 
(which are discrete variables with a domain of 0, 90, 180, 270), and line 
locations and dimensions (which have interval domains represented by a 
min and a max). The constraints are: >, >=, =, +, *. A rectangle imposes 
the following constraints on its variables such as: west line + length = 
east line.
 
Initially the CSP is disconnected. Configurations are formed by placing
design units in geometrical and topological relations, and constraints are
added to the CSP as result. For example, placing two design units adjacent to
each other makes two lines equal. Different topological configurations result
in different constraints being added to the CSP. After the addition of each
new constraint propagation takes place, ensuring global consistency.
Initially, interval domains are from negative infinity to positive infinity.
Constraints remove values from the domains of variables by increasing the
minimum and decreasing the maximum until the minimum is equal to the maximum.
This is a least commitment approach where each solution is an abstraction
in which variables still have a range of values.
 
Using a CSP representation enables determining both the topology of a 
layout and the dimensions of the objects in the configuration in one 
integrated process unlike other spatial planning systems which have
to use a two step approach.
 
Space planning is a search process characterized by very large search spaces.
Constraint directed search provides a basic problem solving methodology
for intelligent CAD by providing a formal method for representing domain
knowledge uniformly as constraints, and by using constraints for efficient
search.
 
Constraints reduce search complexity by opportunistically choosing
the most constrained decision at each point. The knowledge that enables 
us to identify opportunistic decisions in space planning is formulated 
as a certainty measure associated with each constraint.
 
 
67.6YESHERON::ROACHTANSTAAFL !Thu Jun 08 1989 10:18110
Printed by: Pat Roach                                    Document Number: 007534
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-Jun-1989 01:15am ETE
                                        From:     TSS
                                                  TSS@AIADM@ISTG@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ANNOUNCEMENT, 6/14, Birmingham




 

	TITLE:		"Automating the Logical Design 
                        of Computer Hardware"

	SPEAKER:	PROFESSOR WILLIAM BIRMINGHAM
			University of Michigan

	DATE:		Tuesday, June 14, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Conference Room, DLB12

	HOST:		GEORG KLINKER
			Principal Software Engineer
			AI Research Group, AITC
 
	This talk will overview the operation of the MICON system's
	M1 and CGEN.  The results of an extensive set of experiments 
	used to test these tools will also be presented.

	Background Information:
	----------------------

 	Rapid technological advances in micro-processors and related 
	components provide great opportunities for hardware designers 
	to build sophisticated computers.  However, these opportunities 
	come with a price. 

	Computer manufacturers face greater pressure than ever to bring 
	more powerful, but less costly,	machines to market quickly.  
	This, in turn, increases the demands on	hardware designers, as 
	they are expected to deliver what the market requests.  

	Designers face another problem, as they must master a greater 
	amount of technical information than ever before, in order to 
	fully exploit the capabilities of today's components.  M1 is 
	one mechanism for helping designers manage these problems by 
	automating portions of the design task. 
	
	The MICON system is an integrated collection of programs which 
	automatically synthesizes small computer system hardware from 
	high level specifications. 

	The synthesis task is performed by a knowledge-based tool 
	called MICON Synthesizer Version 1 (dubbed M1).  Its power 
	comes from its problem-solving approach and large knowledge-base.

	CGEN is an automated knowledge acquisition tool which allows 
	hardware designers, with no training in artificial intelligence,
 	to deposit their expertise into M1's knowledge-base.






	"Automating the Logical Design
	of Computer Hardware"
	William Birmingham
	Tuesday, June 14, 1988
	Page Two




	CGEN plays a particularly important role by allowing M1's 
	knowledge-base to be easily updated to track technological 
	advances.  Both CGEN and M1 are fully functional, having 
	captured thousands of rules and produced dozens of working 
	designs.

	William Birmingham is currently as assistant professor in the
	Electrical Engineering and Computer Science Dept. at the 
	University of Michigan, where he is also a member of the
	Advanced Computer Architecture Laboratory.

	Prof. Birmingham received a Ph.D. in electrical engineering 
	in 1988, a M.S. in 1983, and a B.S. in 1982 from Carnegie
	Mellon University.

	While at CMU he participated in the development of advanced
	computer-aided design (CAD) tools, utilizing AI techniques,
	and the design of several integrated circuits.
	
	At the University of Michigan he is pursuing research in the
	areas of automated knowledge acquisition and knowledge-based
	design tools, and very high speed computer design.  Prof. 
	Birmingham has spent several years in industry developing
	AI-based CAD tools.  He is a member of IEEE, ACM, and Sigma Xi.
67.7????HERON::ROACHTANSTAAFL !Fri Jun 09 1989 19:4891
Printed by: Pat Roach                                    Document Number: 007563
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     09-Jun-1989 06:46pm ETE
                                        From:     Pat Roach
                                                  ROACH
                                        Dept:     E/ACT Technology
                                        Tel No:   DTN 828-5784

TO:  Florence Bakry                       ( BAKRY )


Subject: ????

    Flo,
    
    I remember that you mentioned that you were interested in DECwindow 
    training. I think that the following video announcement fits your 
    needs. 
    
    Wanna order it for us?
    
    
    pat


                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     09-Jun-1989 06:33pm ETE
                                        From:     TSS
                                                  TSS@AIADM@ISTG@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ANNOUNCEMENT 6/13, Tom Spine

			***********************
			ASDS SEMINAR SERIES #16
			***********************


	TITLE:  	CUSTOMERS USING/DISCUSSING DECwindows SOFTWARE:
				A VIDEOTAPE PRESENTATION

	SPEAKER:	TOM SPINE, SOFTWARE USABILITY ENGINEERING

	TIME:		TUESDAY, JUNE 13, 1989, 2:00 PM

	PLACE:		PACIFIC CONFERENCE ROOM, DLB12

	HOST:		LUCEY BOWEN

	ABSTRACT:

        Between October and December, 1988, the Software Usability Engineering
        group visited six DECwindows field test customer sites, and conducted
        in-depth interviews with 19 DECwindows users.  The customer sites
        included three different McDonnell Douglas sites, New York University,
        Goldman Sachs in New York City, and KMS Fusion in Ann Arbor, Michigan.

        At this seminar I will present a 45-minute videotape which contains
        excerpts from these interviews.  The purpose of the videotape is to
        bring you a systems-level view of the DECwindows user interface, in our
        customers' own words.  The videotape illustrates four themes that
        emerged from the interviews: (1) learning and productivity, (2)
        pointing and clicking with the mouse versus typing, (3) customizing the
        system, and (4) preferences and assessments of the DECwindows system.

        This seminar is intended for anyone who is involved in the development
        of user interfaces for DECwindows applications.

        There will be time for questions and discussion after the videotape.

	
	ABOUT THE SPEAKER: 

        Tom Spine received a B.A. in Mathematics from Washington University in
        St. Louis, and a M.S. in Industrial Engineering from Virginia Tech.  He
        joined Digital in 1985, where the first project he worked on was the
        user interface for VAX Notes.  Tom was a member of the core team that
        designed the user interface for DECwindows, and he was centrally
        involved in the design of several DECwindows application interfaces,
        including the Bookreader, VAX Debug, CMS, DTM, and PCA interfaces. 
        Tom's current responsibilities include acting as project leader for
        Software Usability Engineering's usability consulting services, and
        producing the next version of the XUI Style Guide.
67.8Announcing Three Brown Bag Seminars for June. MarkHERON::ROACHTANSTAAFL !Mon Jun 12 1989 12:2989
Printed by: Pat Roach                                    Document Number: 007566
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     09-Jun-1989 10:37pm ETE
                                        From:     CERVA
                                                  CERVA@ISTG@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Announcing Three Brown Bag Seminars for June.  Mark your calendars.

Announcing three Brown Bag Seminars for the Month of June.  All at noon in
the Leopard Conference Room, DLB5-3/D9 for one hour. Please mark your calendars!
These seminars will be videotaped.

1)  John Comella    Poor Man's Hypertext     Wednesday 14-JUN-1989.

2)  Brad Whitehall  Integrated Learning      Friday 16-JUN-1989.

3)  Chuck Marshall  The Center for Integrated Manufacturing Decision Systems
		    at Carnegie Mellon University     Friday 30-JUN-1989.


1)  John Comella    Poor Man's Hypertext     Wednesday 14-JUN-1989.

Abstract: Customized VMS_HELP can be used with any application because it 
	  (HELP) is callable. 

	  Given a template, it is quite easy to put HELP into your software.  
	  The template is helpful because callable-help has a few minor, but 
	  annoying, idiosyncracies which must be dealt with. 

	  You can set up any number of distinct help librararies so that you 
	  can call a different library on each call.  On the screen it looks 
	  exactly like VMS_HELP, except that you have control over the screen 
	  format and what is available.  Building libraries is simple.

	  I have a template in BASIC, but it is obvious how to run it in LISP, 
	  C or other language (under VMS). 

	  (John Comella is a software engineer in AISG/ISTG.)


2)  Brad Whitehall  Integrated Learning      Friday 16-JUN-1989.

Abstract: An integrated learning technique combines the strengths of analytical 
	  and empirical approaches to machine learning.  The integrated 
	  approach is able to learn in difficult situations, such as computer 
	  aided process planning for manufacturing engineering, because it 
	  does not assume a complete and accurate set of background knowledge.  
	  Yet, it is able to use the knowledge provided, both from rules and 
	  examples to guide the learning process to induce more meaningful 
	  rules.  The capabilities of the system are demonstrated by creating 
	  rules for an expert system that performs process planning for the 
	  manufacturing of machined parts.

	  The goal of the current project is to apply integrated learning 
	  techniques to semiconductor problems.

	  (Brad Whitehall is a PhD student at the University of Illinois.
	   He is working this summer with the AISG/ISTG group.)


3)  Chuck Marshall  The Center for Integrated Manufacturing Decision Systems
		    at Carnegie Mellon University     Friday 30-JUN-1989.

Abstract: The Center for Integrated Manufacturing Decision Systems has been
	  created to conduct research in intelligent decision systems for 
	  engineering and manufacturing problem solving.  The  research
	  agenda strives toward two goals:

	     1) decrease product life cycle times by enabling performance
		of tasks in parallel, by providing better communication
		and integration of decision making.

	     2) increase decision quality at each stage of the product
		manufacturing cycle (design, planning, production, distribution,
		and field service), by utilizing knowledge based systems
		and coordinating decisions.

	  (Chuck Marshall from ASTG/ISTG is working this year at Carnegie
	   Mellon.  He will discuss the center as well as his individual
	   project dealing with getting two different knowledge bases with
	   different structures and models to communicate.)
67.9** AI QUARTERLY COMMUNICATION MEETING ** JUNE 26THHERON::ROACHTANSTAAFL !Wed Jun 14 1989 10:4336
Printed by: Pat Roach                                    Document Number: 007602
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     14-Jun-1989 00:03am ETE
                                        From:     HUNT
                                                  HUNT@AIPERS@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ** AI QUARTERLY COMMUNICATION MEETING **  JUNE 26TH

***** THIS MESSAGE IS FROM DENNIS O'CONNOR AND STEVE SCHUIT *****




Everyone is cordially invited to the next AITC Quarterly meeting on
June 26th from 9:00-11:00am in the DLB5 cafeteria.  The agenda is as
follows:

        Messages that we heard from Jack Smith
                 Dennis O'Connor
                 Member of AITC staff - Steve Gutz
                 Member of AITC       - Cynthia Young

        Research highlights             John McDermott

        Rewards & Recognition           Everyone


We expect full attendance at this important communication meeting.
67.10** SEMINAR ** MANAGEMENT EXCELLENCE: A MODEL FOR SHERON::ROACHTANSTAAFL !Wed Jun 14 1989 19:4957
Printed by: Pat Roach                                    Document Number: 007618
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     14-Jun-1989 06:16pm ETE
                                        From:     HUNT
                                                  HUNT@AIPERS@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ** SEMINAR ** MANAGEMENT EXCELLENCE:  A MODEL FOR SUCCESS

                           BROWN BAG LUNCH-TIME


TOPIC:    Management Excellence:  A Model for Success.

This first Brown Bag Seminar will focus on understanding and increasing
your management effectiveness.

During this session you will:

-  Examine your assumptions about managerial effectiveness.

-  Learn about effective and ineffective managerial experiences.

-  Leave with a model to guide organizational transitions.

-  Learn where to get more information on this and other related topics.

Join us as we examine some of the variables and hypothesis which
influence managerial effectiveness.


DATE:   Friday, July 14th, 1989

PLACE:  Atlantic Conference Room, DLB12

TIME:   12 noon to 1:00 pm

DISCUSSION
LEADER:    Michael Halperin, president of Halperin and Associates,
           is a member of the Distributed Systems/Digital Management
           Education faculty Team.  He has worked as a consultant
           and trained for thirteen years, designing and delivering
           management education programs and consulting to clients
           on team building and organizational issues.  Mr. Halperin
           has been associated with Digital Management Education for
           eight and a half years where he has managed a variety of
           consulting projects.

If you would like more information on the program, please contact
John LeBlanc at 291-8478.
67.11YESHERON::ROACHTANSTAAFL !Wed Jun 14 1989 21:0452
Printed by: Pat Roach                                    Document Number: 007620
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     14-Jun-1989 07:57pm ETE
                                        From:     TSS
                                                  TSS@AIADM@ISTG@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: 6/15-MR4-3:30PM-R. FRANK

	TITLE: 		"The Computer-Aided Engineering Network 
	                at the Univer. of Michigan"

	SPEAKER:	Randall L. Frank
			Director of Information Technology
			College of Engineering

	DATE:		Thursday, June 15, 1989

	TIME:		3:30-5:00 P.M.

	PLACE:		MRO4 Amphitheatre
	 
	The University of Michigan, College of Engineering, has 
	undertaken a massive program of workstation and network 
	deployment. 

	The goal is to fundamentally change the way that engineering 
	instruction and research are performed.

	This talk will describe the environment in place at Michigan, 
	and will discuss anticipated trends over the next few years.

  	The particular environment at Michigan has been tailored 
	to provide support for the premiere engineering applications 
	in use within the industrial community, tailored, as appropriate, 
	for a university setting.  

	Fundamental to this approach has been the building of an 
	environment that supports shared access to information 
	across a highly heterogeneous computing environment, from 
	Macs and PCs to large scale computers.  Major components 
	on both the undergraduate and graduate curriculum have now 
	been converted to use this environment, with fundamental 
	changes taking place in both the electrical and mechanical 
	computer-aided design curriculum.
67.128/1, DR. BONAR, 9:30AMHERON::ROACHTANSTAAFL !Thu Jul 13 1989 09:2598
Printed by: Pat Roach                                    Document Number: 007974
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Jul-1989 06:55pm ETE
                                        From:     TSS
                                                  TSS@AIADM@ISTG@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: 8/1, DR. BONAR, 9:30AM





	TITLE:		"What Good Are Graphic User Interfaces?, or,
			Beyond the Desktop Cliche"

	SPEAKER:	DR. JEFFREY BONAR
			Virtual Machine Corp. President

	DATE:		Tuesday, August 1, 1989

	TIME:		9:30-11:00 AM

	PLACE:		Pacific Room, DLB12

	HOST:		ED FISHER
			Software Engineering Manager
			SELECT Group, AI Technology Center

	This talk will focus on how graphic user interfaces can and
	should go beyond a desktop.

	Current development of graphical user interfaces has focused
	on refinement of the desktop metaphor.  Unfortunately, the
	desktop metaphor is only one of many possible forms that a
	graphic user interface could take.  

	Marshall McLuhan said that a new medium always begins by 
	imitating the old medium:  cow paths were paved to make roads 
	for the "horseless carriage"; film began by putting a camera 
	in front of a play; and now, computer screens look like desktops.

  	How should an application present itself inside the standardized 
	windows and menus?  How can we best exploit low-level standards 
	like X Window?  How can we seriously grapple with the need for 
	systems that are both easier to learn and more powerful to use?
  
	To answer these questions there will be discussion and illustra-
	tion of three design approaches that lead to more powerful graphic 
	user interfaces: 

		o Building computational capabilities (intelligence)
                  into the screen objects

		o Selecting appropriate screen metaphors based on a
		  cognitive task analysis

		o Exploiting the strengths of human visual perception

	            TM
	VisualACTIVE,  a graphic user interface development environment
	available from Virtual Machine Corp. is presented to illustrate
	the effective use of these approaches.







	"What Good Are Graphic User Interfaces?, or,
	Beyond the Desktop Cliche"
	Tuesday, August 1, 1989
	Page Two



	Jeffrey Bonar received B.A. degrees in Biology and Information 
	Science from the Univ. of CA at Santa Cruz in 1977.  He received
	his M.S. and Ph.D. in Computer and Information Science from 
	U.MASS., Amherst, in 1980 and 1985, respectively.

	In 1984 he joined the Learning Research & Dev. Center at the
	Univ. of Pittsburgh to apply AI technology to education.  He
	has an appointment with the Computer Science Dept. and is a 
	founding member of the I.S. Studies Program.

	Bonar left the Univ. of Pittsburgh in 1988 to pursue commercial
	development of his research in human-computer interaction.  He
	is currently the president of Virtual Machine Corp., marketing
	VisualACTIVE (TM).
67.13Repeating course announcement for Testing KnowledgeHERON::ROACHTANSTAAFL !Thu Jul 13 1989 19:4195
Printed by: Pat Roach                                    Document Number: 007995
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Jul-1989 06:36pm ETE
                                        From:     KING
                                                  KING@AIADM@ISTG@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Repeating course announcement for Testing Knowledge-Based Systems

	Due to low enrollment it is necessary to resend this announcement
	in the hopes of filling the class.

	Please let me know if you are interested in this course.



********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		July 25th and 26th, 1989.

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Michelle Fineblum, Instructor and Course Developer

AUDIENCE and 	  - DEC employees only -
PREREQUISITES:	
		The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1500.

TO REGISTER:	Contact Louise King, AIADM::KING, DTN: 291-8600 with badge
		number, DTN, cost center, group/organization name and manager's 
		name.

CANCELLATION:   10 Days prior to the start date of the course.
67.14Repeating course announcement for Testing KnowledgeHERON::ROACHTANSTAAFL !Mon Jul 17 1989 11:4695
Printed by: Pat Roach                                    Document Number: 007999
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     14-Jul-1989 01:11am ETE
                                        From:     KING
                                                  KING@AIADM@ISTG@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Repeating course announcement for Testing Knowledge-Based Systems

	Due to low enrollment it is necessary to resend this announcement
	in the hopes of filling the class.

	Please let me know if you are interested in this course.



********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		July 25th and 26th, 1989.

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Michelle Fineblum, Instructor and Course Developer

AUDIENCE and 	  - DEC employees only -
PREREQUISITES:	
		The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1500.

TO REGISTER:	Contact Louise King, AIADM::KING, DTN: 291-8600 with badge
		number, DTN, cost center, group/organization name and manager's 
		name.

CANCELLATION:   10 Days prior to the start date of the course.
67.15REMINDER, 7/19, AMPIIHERON::ROACHTANSTAAFL !Wed Jul 19 1989 10:45120
Printed by: Pat Roach                                    Document Number: 008048
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Jul-1989 08:13pm ETE
                                        From:     TSS
                                                  TSS@AIADM@ISTG@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: REMINDER, 7/19, AMPII







	TITLE:		"New CGI/Digital Co-Developed 
		        Modules for Knowledge Craft"

	SPEAKER:	CINDY COSIC
			Carnegie Group, Inc.

	DATE:	 	Wednesday, July 19, 1989

	TIME:		9:30-11:30 AM

	PLACE:		Pacific Room, DLB12

	HOST:		FRANK LYNCH
			Group Engineering Manager
			Advanced Systems & Tools Group

	This seminar will cover MODELPAK and TIMEPAK, modelling and temporal
	representation and reasoning components available with Knowledge
	Craft.  The two new components, Modelpak and Timepak, can be used
	separately or in conjunction with each other or other Knowledge Craft
	components.

	The first component, Modelpak, allows the user to model a discrete
	simulation process.  The package allows the user to vary the con-
	straints and parameters that define the model.  The user can then
	experiment with the process being represented to develop the optimum
	configuration of its components.

	Modelpak allows the user to define and execute a model of events 
	represented as activities and resources.  In Modelpak, an activity
	represents an action that is to occur during the process.  An
	activity may be a single action or it may be a collection of
	activities referred to as a complex-activity.  Resources are the
	physical substance typically required for an activity's execution.
	Resources have capacities that may or may not diminish with use,
	depending on their nature.

	Once the user has defined a model's activities and resources, and
	their respective characteristics, Modelpak allows the user to link 
	the activities and resources to parallel the manner in which they
	are used.  The user attaches the resources to any activities by
	which they are used, and links the activities themselves to form
	a sequence of actions.  The result is a hierarchical model of
	activities and resources that combine to represent a discrete 
	process.

	The second component, Timepak, provides a foundation upon which 
	users can represent and reason about temporal occurrences.  Timepak 
	consists of two primary subsystems:  absolute time and relative time.







	"New CGI/Digital Co-Developed Modules for Knowledge Craft"
	Wednesday, July 19, 1988
	Page Two





	Absolute time allows the user to create a temporal representation
	associated with fixed or definite occurrences.  The user first creates
	units of time to be used in the representation.  The user can then
	assign these units to specific points in time and define extended
	intervals bounded by the points.  Timepak's absolute time repre-
	sentation also accommodates definite spans of time not associated
	with any particular starting or ending points.

	Relative time allows the user to define events not constrained by
	specific time references.  The user groups these events and creates
	time-relations between the members of a group and, possibly,
	between events from different groups.  In Timepak, time-relations
	consist of statements such as "one event occurs BEFORE another."
	Once a network of events has been defined, grouped, and related, 
	the user can determine the relationships that exist between any 
	of the events in the network, drawing conclusions and reasoning
	about the process represented by the events.

	Both Modelpak and Timepak are built using object-oriented programming
	(OOP).  The elements of each package are represented as objects. Each
	object consists of a schema, with mothods that define its behavior.
	An object's functionality is accessed via messages sent to its 
	methods.  Modelpak and Timepak are also both implemented in CRL, 
	enabling all CRL functions to be used with either package.

	Cindy Cosic graduated from the Univ. of Pittsburgh in 1984 with
	an M.S., and is currently pursuing a Ph.D. there in Information
	Science on a part-time basis.  She obtained a B.S. from the Univ.
	of Pittsburgh in Physical Therapy in 1981.  Cindy joined the 
	Carnegie Group in July of 1988.

	Other CGI Development Team members include:  Gary Strohm, James
	Bycura, Jeff Kaminski, and Joyce E. Kepner.  (Biographies available
	upon request.)
67.162 Brown Bag seminars on Symbolic Modeling and DiffuHERON::ROACHTANSTAAFL !Mon Jul 24 1989 10:34190
Printed by: Pat Roach                                    Document Number: 008114
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Jul-1989 10:06pm ETE
                                        From:     CERVA
                                                  CERVA@ISTG@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: 2 Brown Bag seminars on Symbolic Modeling and Diffusion of Innovation coming up.  Please mark your calendars  (again).  Tom

Dear AITC,

We are very pleased to announce two important Brown Bag Seminars on 31-JUL
and 16-AUG in the Leopard CR, DLB5-3/D9 at noon.

Kathy Rohulich of AISG will speak 31-JUL-89 on Reality-Based Information 
Systems Modeling, a breakthrough method of combining the symbolic features 
of static diagramming techniques with the dynamic capabilities of discrete-
event simulation techniques to develop complex information systems.

Please consider attending if you wish to learn about a leading-edge method
of planning, designing, and implementing complex information systems at
Digital.

Bill Penney of Business Planning Systems will speak 16-AUG-1989 on Diffusion
of Innovation, the change process and adaption of innovation throughout an
organization.

Please consider attending if you wish to gain insight into how it is possible
to introduce change and innovation (your projects, perhaps?) into Digital.

Also, please don't forget about the three Brown Bag Seminars taking place 
on 25-JUL-1989 in Leopard CR, DLB5-3/D9 at 9:00, Noon, and 1:30.  I have 
appended a copy of the summary announcement for these three seminars.

Thank you,
Tom Cerva


1)  REALITY-BASED INFORMATION SYSTEMS MODELING

    WHO:       Kathy Rohulich, AISG
    WHEN:      Noon to 1 PM, 31-JUL-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      Tom Cerva, AISG
    FOLLOW-UP: Kathy's DTN is 291-8485 and she is at DLB5-3/D5.

2)  DIFFUSION OF INNOVATION

    WHO:       Bill Penney, Business Planning Systems
    WHEN:      Noon to 1 PM, 16-AUG-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      Mary Lewis, AISG
    FOLLOW-UP: Please contact Bill Penney directly or through Mary Lewis.


NOTE:  These two talks will be videotaped.  

NOTE:  Please consider attending the three talks on 25-JUL (details at end).



1)  REALITY-BASED INFORMATION SYSTEMS MODELING

    WHO:       Kathy Rohulich, AISG
    WHEN:      Noon to 1 PM, 31-JUL-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      Tom Cerva, AISG
    FOLLOW-UP: Kathy's DTN is 291-8485 and she is at DLB5-3/D5.


 REALITY-BASED INFORMATION SYSTEMS MODELING
     

 In recent years information systems have evolved from back-room
 administration and accounting tools into more corporate and strategic
 roles, that are essential to improving an organization's competitive
 edge. These 'enterprise' systems fully integrate all processes and
 systems related to a business thereby, requiring maximum coordination
 and integration.
             
 The risks associated with developing these complex information systems
 can range from time-cost overruns to project failure. The degree of
 risk associated with the development of these systems bears a strong
 relationship to the presence and extent of certain development factors
 which include:  1) the size and number of interactions of the
 development process (e.g., the number of departments involved),  2) the
 technical complexity (e.g., a conventional local batch system with a 24
 hour turn-around, or a more demanding geographically distributed system
 with a few seconds response time) and, 3) the complexity of the
 business application system (e.g., a system with centralized control,
 or an object-oriented system design with cooperative problem solvers)
                                     
 Several methodologies exist that can be used to reduce the risk during
 the first phase of complex information system development which include
 static diagramming techniques (IDEF and SADT) and conventional
 simulation techniques. Static diagramming techniques lack dynamic
 capabilities and conventional simulation techniques focus on numerical
 rather than symbolic abstraction.
                      
 This talk will describe a computer-aided systems engineering (CASE)
 methodology which couples the symbolic features of static diagramming
 techniques with the dynamic capabilities of discrete-event simulation
 techniques. The talk will also cover how this methodology is being used
 at Digital today to design and analyze complex information systems in
 key business areas such as order processing, field service and
 manufacturing.


2)  DIFFUSION OF INNOVATION

    WHO:       Bill Penney, Business Planning Systems
    WHEN:      Noon to 1 PM, 16-AUG-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      Mary Lewis, AISG
    FOLLOW-UP: Please contact Bill Penney directly or through Mary Lewis.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

                          Diffusion of Innovation
                          =======================

    High tech/high touch is effective use of technology by people.  The
    change process and adaption of innovation is a human one, not a
    technological one.  Human population innovativeness characteristics are
    discussed, a process model of innovation diffusion is supplied, and the
    social backdrop for change is reviewed.  A checklist for managing change is
    supplied at the conclusion of the talk.
    
                             About Bill Penney
                             =================
    
    Bill Penney is an industrial engineer with a lifelong interest around
    helping people effectively utilize computing technologies.  He marketed
    commercial timesharing services, was an independent computer
    consultant, and published a newsletter on management of small business
    computing.  At DEC, he authored TECHNOCHANGE:  MANAGEMENT OF END-USER
    COMPUTING.  Currently, he's an IS consultant in Business Planning
    Systems.
     


Copy of previous Brown Bag announcements.

ANNOUNCING THREE BROWN BAG TALKS ON 25-JUL-1989 in DLB5,
Artificial Intelligence Technology Center, Marlboro, Mass.

All Three Talks in Leopard CR, DLB5-3/D9, at 9:00, 12:00, and 13:30.  Please
mark your calendars for this day.



1)  SHOPTALK, An Integrated Interface for Decision Support in Manufacturing

    WHO:       Philip R. Cohen, Artificial Intelligence Center,
	       SRI International
    WHEN:      9:00 AM to 10:15, 25-JUL-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      Stew Hoover, AISG
    FOLLOW-UP: We have scheduled the Cheetah CR, DLB5-3/C9, from 10:30 to
		noon for follow-up conversation with Dr. Cohen


2)  Introduction to Neural Nets 

    WHO:       Kaihu Chen, AISG
	       Professor Oded Maimon, University of Tel Aviv
    WHERE:     Leopard CR, DLB5-3/D9
    WHEN:      Noon to 1:15 PM, 25-JUL-1989
    HOST:      Henry Teng, AISG
    FOLLOW-UP: Kaihu Chen (291-8489) is a fulltime AISG Digital employee and 
	       Oded Maimon (291-8283) is working with AISG for the summer.


3)  An Overview of the Stanford Research Program in Manufacturing Science 
    and Technology for VLSI

    WHO:       Ernest J. Wood, Center for Integrated Systems, 
	       Stanford University
    WHEN:      1:30 PM to 2:45 PM, 25-JUL-1989
    WHERE:     Leopard CR, DLB5-3/D9
    HOST:      David Cavanaugh, AISG
    FOLLOW-UP: We have scheduled the Cheetah CR, DLB5-3/C9, from 3:00 to
		4:30 for follow-up conversation with Dr. Wood
67.17YESHERON::ROACHTANSTAAFL !Wed Jul 26 1989 10:3353
Printed by: Pat Roach                                    Document Number: 008166
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     25-Jul-1989 08:01pm ETE
                                        From:     SOOHO
                                                  SOOHO@AITG@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: VAX LISP V3.0 TSS being held in ZKO on 8/1/89

From:	WECARE::ZKO_TRAINING "25-Jul-1989 1013" 25-JUL-1989 10:20:05.97
To:	@TSS.DIS;
CC:	
Subj:	ZKO TSS - "A Multi-threaded Interactive DECwindows Programming...."

Spit Brook Technical Seminars


Title:    "A Multi-Threaded Interactive DECWindows Programming Environment"

Speakers: Walter van Roggen, Greg Mangan, Richard Wells 
          Digital -  AITG

Date:	  Tuesday, August 1, 1989
Time:     10:00 - 11:00 am
Place:    Babbage Auditorium, ZKO-1


As software systems become ever more complex, designing, testing,
and modifying such systems requires increasingly sophisticated
interactive development environments.  These environments feature
window-based user-interface and code-development tools, multi-thread
capability, automatic memory management, object-oriented programming,
incremental compilation, and a simple, flexible, extensible syntax.

VAX LISP V3.0 provides these features.  We present the design of some
of these features in the context of the DECWindows window system and
our Common Multithread Architecture inspired multi-thread facility.
We also present some of the difficulties we faced in our implementation,
and we suggest some DECWindows changes that would benefit future
programming environments.

Technical Sponsor :  Bob Conti - TLE

                      - no reservations required -

THE VIDEOTAPE OF THIS SEMINAR WILL BE AVAILABLE IN THE SPIT BROOK LIBRARY
67.18Upcoming Course - Testing Knowledge-Based SystemsHERON::ROACHTANSTAAFL !Fri Aug 11 1989 18:5191
Printed by: Pat Roach                                    Document Number: 008403
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Aug-1989 05:14pm ETE
                                        From:     ZIESEMANN
                                                  ZIESEMANN@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Upcoming Course - Testing Knowledge-Based Systems


******************* From Louise King AI TRAINING GROUP *************************


********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		September 21 - 22nd, 1989.

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Theresa Chow, Instructor

AUDIENCE and
PREREQUISITES:	The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1500.

TO REGISTER:	Contact Louise King, AIADM::KING, DTN: 291-8600 with badge
		number, DTN, cost center, group/organization name and manager's 
		name.

CANCELLATION:   10 Days prior to the start date of the course.
67.19Upcoming Course - Integrating Knowledge-Based SysteHERON::ROACHTANSTAAFL !Fri Aug 11 1989 18:5287
Printed by: Pat Roach                                    Document Number: 008404
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Aug-1989 05:30pm ETE
                                        From:     ZIESEMANN
                                                  ZIESEMANN@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Upcoming Course - Integrating Knowledge-Based Systems


*************** From Louise King  AI TRAINING GROUP ****************************


********************************************************************************
	           INTEGRATING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

       DATE:	September 18 - 20th, 1989

   LOCATION:	DLB12-1 
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:    The successful delivery of an application that uses artificial
                intelligence (AI) tools and techniques frequently depends
                on the ability of the system to coexist with other
                applications and existing data.  Using examples and
                exercises from case studies, this course presents
                guidelines and methods for integrating knowledge-based or
                expert systems into the application environment and easing
                a system's transition from prototype to production.  The
                course will highlight issues in project planning, system
                life cycle, knowledge representation, and implementation
                that must be addressed to successfully integrate a
                knowledge-based system into a business application
                portfolio.

      GOALS:    Upon completion of the course, participants will be able
                to:

                    - Define integration requirements for knowledge-based
                      applications

                    - Incorporate integration issues into the project plan

                    - Define the relationship (and a transformation)
                      between a relational data definition and a
                      knowledge representation

                    - Write simple VAX LISP programs that access non-LISP
                      routines and/or data files

                    - Identify trends toward the integration of databases
                      and AI applications
 
 PRESENTER:	John Arnold

  AUDIENCE:     Application developers, programmers, or software engineers
                involved with the delivery of knowledge-based or other AI
                systems. 
                Project leaders or technical managers of applications that
                have or are considering a knowledge-based component.


LENGTH & FORMAT:    3 days, Lecture/Lab


PREREQUISITES:  - Completion of the Expert Systems Training Program or
                  equivalent

                - Applications development experience

                - Familiarity with LISP programming

        COST:   $2,300.

    REGISTER:	Send mail to AIADM::KING.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.20REMINDER, Drs. Cohen & Lee, DLB12, 8/30, 9:30AMHERON::ROACHTANSTAAFL !Tue Aug 29 1989 18:53102
Printed by: Pat Roach                                    Document Number: 008568
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     29-Aug-1989 05:40pm ETE
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: REMINDER, Drs. Cohen & Lee, DLB12, 8/30, 9:30AM







	TITLE:		"A Multi-Echelon Inventory System 
			for Service Logistics Management"

	SPEAKERS:	DR. MORRIS A. COHEN, University of Pennsylvania
		  and	DR. HAU L. LEE, Stanford University

	DATE:		Wednesday, August 30, 1989

	TIME:		9:30-11:00 AM

	PLACE:		Pacific Room, DLB12

	HOSTS:		MITCH TSENG, Manager 
		  and	PHIL SOO, Senior Software Engineer
			Applied Intelligent Systems Group
			AI Technology Center


	IBM has recently implemented a system for flexible, optimal control
	of spare parts inventory developed for its U.S. after-sales network.

	Professors Morris A. Cohen and Hau L. Lee are the creators of the
	theories and methodologies embeddeed in the IBM system.  (They have
	also applied their methodology to automobile spare parts logistics
	operations.) In this talk, they will describe their new methodology
	and disucss its development and implementation.  

	The system utilizes state-of-the-art methodologies for spare parts
	multi-echelon inventory management and utilization forecasting.  It
	provides stocking decisions for over 15,000 stocking locations and
	supports 200,000 part numbers and 1,000 active products.  

	Since its implementation, a 20-25% reduction in inventory investment, 
	or in excess of a QUARTER OF A BILLION DOLLARS at selling price, has 
	been achieved.  Finally, the system also saves $20 million in the 
	annual operating costs for the service logistics system. 

	Dr. Morris A. Cohen is Prof. of Decision Sciences at the Wharton 
	School, University of PA.  He holds a B.A.Sc. in engineering 
	sciences from the Univ. or Toronto, an M.S. in industrial engineer-
	ing and	management science and a Ph.D. in operations research from 
	Northwestern Univ.

	Dr. Hau L. Lee is Associate Prof. of Industrial Engineering and
	Engineering Management at Stanford Univ.  He holds a B.S. in 
	economics and statistics from the Univ. of Hong Kong, an M.S.
	in operations research from the London School of Economics and
	a Ph.D. in operations research from the Wharton School of the
	Univ. of PA.







	"A Multi-Echelong Inventory System
	for Service Logistics Management"
	August 30, 1989
	Page Two




	Professors Cohen and Lee have been actively involved in research
	and consulting in the areas of manufacturing strategy, inventory
	control, logistics, distribution, technology management, quality
	assurance and control and production planning for various
	organizations in both the public and private sectors.  Their 
	joint research on logistics has received a runner-up award in 
	the 1989 Edelman prize competition for the best practice in
	Management Science (sponsored by the Institute for Management
	Science).  Their research on global manufacturing was awarded
	the 1989 Lauder Institute/TIMS prize for the best advances in
	the theory and practice of international management science.
	


	NOTE:  As a courtesy to the speakers, we ask attendeed to arrive
	       prior to scheduled start time.  Thank you.
67.21CALENDAR UPDATEHERON::ROACHTANSTAAFL !Tue Sep 12 1989 09:1945
Printed by: Pat Roach                                    Document Number: 008699
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Sep-1989 01:51am ETE
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: CALENDAR UPDATE





		AI TECHNOLOGY CENTER TECHNICAL SEMINAR SERIES

			     C.A.L.E.N.D.E.R. 

			
				SEPTEMBER
			        ---------


DATE/TIME/PLACE		SPEAKER/AFFILIATION	TITLE
---------------         -------------------	-----

9/20, Wednesday		Walter Van Roggen,  	"A MultI-Threaded Interactive
9:30-11:00AM		Greg Mangan		DECWindows Programming Environ-
DLB12, Pacific Room	Richard Wells		ment"
			Digital, AITG

9/22, Friday		Dr. William C. Hill	"The Mind at AI:  Horseless
9:30-11:00AM		MCC			Carriage to Clock"
DLB5, Cafeteria

9/26, Tuesday		Lee Peterson		"Graduate Engineering Education"
9:30-11:30AM		Digital			(GEEP)
DLB12, Pacific Room
	
67.23EPITOOL COURSE OFFERINGHERON::ROACHTANSTAAFL !Sat Sep 16 1989 15:57128
Printed by: Pat Roach                                    Document Number: 008756
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     16-Sep-1989 03:10am ETE
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL COURSE OFFERING


             ******************************************************
             *                                                    *
             *       Educational Services Intelligent Systems     *
             *                                                    *
	     *		      is happy to announce                *
             *                                                    *
             *                1 st offering of the                *            
             *                                                    *
	     *			EPITOOL course                    *
             *                                                    *
             *         (open to Digital Employees only)           *
             ******************************************************

	
Dates:		17 - 20 October 1989 

Location:	Carribean Classroom in DLB12,
		295 Donald Lynch Boulevard
		Marlboro

Price:          $ 595/person

Registration:	Please have your manager send a message to Robyn Maki, 
		(DTN 291-8893) at ESIS::MAKI including the following
		information:
		
		- Name of person attending the course
		- Email, phone and mailstop
		- Cost center

		Cost center manager needs also to send a message to
		Robyn Maki with approval. 

		Number of attendants are limited to 20.


Information:	For more information contact 
		Malin Goodwin at ESIS::GOODWIN or DTN 291-8894
		

About EPITOOL:
--------------
	EPITOOL is a Lisp based hybrid expert system tool developed by 
	Epitec AB in Link�ping, Sweden. As announced at IJCAI in Detroit,
	EPITOOL will be a software product distributed and sold by Digital.

	EPITOOL	has a wide range of knowledge representations including 
	forward and backward chaining rules, objects and methods, and 
	functions for represenation of procedural knowledge. Epitool also
	provides functions for creation of Decwindows-style application
	graphics, and call-in  call-out to any VAX supported language.

	Development of Knowledge bases can either be done on a VAX station,
	using development graphics, windows and menus, or on a regular
	terminal using Epitools own command language.  
	
        
About the course:
-----------------

		--------------------------------------
        Observe: This is a unique opportunity to get your EPITOOL training
	from two representants of Epitec, Sweden.
                --------------------------------------
	        
	The course consists of lectures and labs. Labs will provide the 
	students with a hands on experience working with the tool. The
	topics covered in the course are:

	- Knowledge Representation in Epitool: 
		Concepts, Aspects, Individuals, Objects-Methods.

	- Actions:
	  	Rules, Backward and Forward chaining. Functions.
		Rule sets.

	- Development tools:
		Error handling, Debugging

	- Values:
		Uncertain, Unknown, Descriptions, Groups.

	- End user interface:
		Questions, Application graphics, Explanations,
		Streams and active paths

	- Integration

	- Customizing the development environment.
		
	



About the instructors:
----------------------


	Kristina Ernstsson - Educational Training Manager at Epitec AB

	Kristina has been with Epitec for 2 years and is responsible for
	the Epitecs Epitool training. She was earlier in the group for
	Programming Environments at Link�ping University, Sweden. 



	Bj�rn M�ller - Project Manager Knowledge Systems at Epitec AB

	Bj�rn has 4 years of experience with Epitool and consulting, and 
	has been with Epitec since the founding of the company. He has 
	been involved in several financial applications.
67.24EPITOOL COURSE OFFERINGHERON::ROACHTANSTAAFL !Sat Sep 16 1989 15:58128
Printed by: Pat Roach                                    Document Number: 008756
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     16-Sep-1989 03:10am ETE
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL COURSE OFFERING


             ******************************************************
             *                                                    *
             *       Educational Services Intelligent Systems     *
             *                                                    *
	     *		      is happy to announce                *
             *                                                    *
             *                1 st offering of the                *            
             *                                                    *
	     *			EPITOOL course                    *
             *                                                    *
             *         (open to Digital Employees only)           *
             ******************************************************

	
Dates:		17 - 20 October 1989 

Location:	Carribean Classroom in DLB12,
		295 Donald Lynch Boulevard
		Marlboro

Price:          $ 595/person

Registration:	Please have your manager send a message to Robyn Maki, 
		(DTN 291-8893) at ESIS::MAKI including the following
		information:
		
		- Name of person attending the course
		- Email, phone and mailstop
		- Cost center

		Cost center manager needs also to send a message to
		Robyn Maki with approval. 

		Number of attendants are limited to 20.


Information:	For more information contact 
		Malin Goodwin at ESIS::GOODWIN or DTN 291-8894
		

About EPITOOL:
--------------
	EPITOOL is a Lisp based hybrid expert system tool developed by 
	Epitec AB in Link�ping, Sweden. As announced at IJCAI in Detroit,
	EPITOOL will be a software product distributed and sold by Digital.

	EPITOOL	has a wide range of knowledge representations including 
	forward and backward chaining rules, objects and methods, and 
	functions for represenation of procedural knowledge. Epitool also
	provides functions for creation of Decwindows-style application
	graphics, and call-in  call-out to any VAX supported language.

	Development of Knowledge bases can either be done on a VAX station,
	using development graphics, windows and menus, or on a regular
	terminal using Epitools own command language.  
	
        
About the course:
-----------------

		--------------------------------------
        Observe: This is a unique opportunity to get your EPITOOL training
	from two representants of Epitec, Sweden.
                --------------------------------------
	        
	The course consists of lectures and labs. Labs will provide the 
	students with a hands on experience working with the tool. The
	topics covered in the course are:

	- Knowledge Representation in Epitool: 
		Concepts, Aspects, Individuals, Objects-Methods.

	- Actions:
	  	Rules, Backward and Forward chaining. Functions.
		Rule sets.

	- Development tools:
		Error handling, Debugging

	- Values:
		Uncertain, Unknown, Descriptions, Groups.

	- End user interface:
		Questions, Application graphics, Explanations,
		Streams and active paths

	- Integration

	- Customizing the development environment.
		
	



About the instructors:
----------------------


	Kristina Ernstsson - Educational Training Manager at Epitec AB

	Kristina has been with Epitec for 2 years and is responsible for
	the Epitecs Epitool training. She was earlier in the group for
	Programming Environments at Link�ping University, Sweden. 



	Bj�rn M�ller - Project Manager Knowledge Systems at Epitec AB

	Bj�rn has 4 years of experience with Epitool and consulting, and 
	has been with Epitec since the founding of the company. He has 
	been involved in several financial applications.
67.25EPITOOL DEMO, 9/26HERON::ROACHTANSTAAFL !Mon Sep 25 1989 11:5860
Printed by: Pat Roach                                    Document Number: 008788
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Sep-1989 07:28pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL DEMO, 9/26


	TITLE:		Epitool Presentation and Demonstration

	SPEAKER:	PERTTI JAUHIAINEN
		        U.S. Product Support, Epitec
	               
	DATE:		Tuesday, September 26, 1989

	TIME and	
	LOCATION:     	12 Noon - 1:00PM - Presentation	
			Leopard Conf. Room, DLB5-3/E9

			1:00-2:00PM - Demonstration
			DLB5 First Floor Demo Room
	

Pertti Jauhiainen is the Product Support man in U.S for Epitec.  He will 
also do presentations and demonstrations of Epitool. Support office is 
located in Framingham, Ma, and our headquaters in Linkoping, Sweden.

To use Epitool, you order the software via SDC (Software Distribution Center),
and you have to only pay the maintenance costs. 

Abstract: 

Epitool is an expert system development tool that provides software engineers
with a rich, hybrid knowledge representation language. While combining the
power of object-oriented, rule-based and procedural problem solving, the tool
is designed to be easy to use for software engineers inexperienced in building
expert systems. Epitool also provides a number of features to facilitate the
delivery of applications, such as end-user graphics, explanation facilities,
and an optimized execution environment that allows developers to build small 
systems.

Both devlopment and delivery environment are available for Epitool. Epitool
Development Environment (Epitool/DE) provides the software engineer with
advanced aids including debugging and analysis services, specialized editors,
and a variety of object browsers. Although Epitool can be used for executing
an application in a production environment, the environment is designed for
the development phase of a project.

The Epitool Execution Environment (Epitool/EE) is optimized for efficient
execution of completed applications. In addition, shareable images can be built
for delivery.
67.26LIPKA, 10/12, 9:30AMHERON::ROACHTANSTAAFL !Mon Sep 25 1989 12:03119
Printed by: Pat Roach                                    Document Number: 008796
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Sep-1989 10:20pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: LIPKA, 10/12, 9:30AM







	TITLE: 		"Using Models for Engineering Systems"

	SPEAKER:	DR. STEPHEN LIPKA
		        Advanced Systems and Tools Group
			AI Technology Center

	DATE:		Thursday, October 12, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room, DLB12

	HOST:		CHRIS CHERPAS
			AI Training and Education Group
			AI Technology Center

	
	This talk will present a limited set of model types, including 
	IDEF0 (functional) modeling and Entity-Relationship-Attribute 
	(information) modeling. Their capabilities will be presented, 
	and their range of usefulness will be discussed. 

	An example drawn from a software prototype developed within 
	Digital will illustrate their use. A work-flow model developed 
	for that prototype will also be presented, and the encoding of 
	that model into a work-flow manager will be briefly discussed.
	

	Background:
	----------

	If one were to characterize the spectrum of single-person 
	software development efforts, one would likely find that the 
	developer was knowledgeable either about the domain being 
	supported or about software. 

	The best of both worlds occurs when a software-knowledgeable 
	developer is adequately knowledgeable about the domain being 
	supported.  This usually occurs when the developer creates a 
	model (perhaps mental) of the domain. 

	For larger projects to succeed, it is also necessary that all 
	members within any one group (domain-knowledgeable or software-
	knowledgeable) agree, i.e., have the same model in mind. 

	Formal models are a useful tool in developing software systems. 
	Useful models can describe (business or software) "systems" 
	rigorously, to some well understood tolerance. Such models are 
	useful for conveying understanding within and across teams. 







	"Using Models for Engineering Systems"
	Dr. Stephen Lipka, October 12, 1989
	Page Two



	
	Models provide visibility into the learning or design process 
	and provide a basis for organized discussions. Well-chosen 
	models make it easier to judge the quality of a "system" when 
	judging by the appropriate design rules. And some models may be 
	encoded in such a way as to drive execution of a system or allow 
	machine-based reasoning.

	Throughout the system development process, a wide spectrum of 
	models can be profitably applied. Some are more useful for 
	business.  Some are appropriate for detailed software.  But, 
	few are appropriate for a wide range of purposes. 

	The important issues in choosing a model type, therefore, are 
	understanding what problem to solve, understanding what a given 
	type of model can do, and understanding how a given model type 
	can be usefully applied. 
	
	Biography:
	---------

	Stephen Lipka is an independent consultant.  He is interested in 
	the use of static and dynamic models for software engineering, 
	multiple-model techniques, and CASE tools that integrate across 
	models. 

	He received a Bachelor's Degree in Physics and a Master of 
	Science in Electrical Engineering at the Polytechnic Institute 
	of Brooklyn. He received his Ph.D. in Computer Science from the 
	State University of New York at Stony Brook in 1975. 

	Stephen is currently working with the Applied Systems and Tools 
	Group at the AI Technology Center in Marlborough, MA. He has a 
	large scope of experience in applying and teaching modeling and 
	software engineering techniques. 
67.27rfd aaTICIPATION IN "GRAD SCHOOL" - PLEASE LET ME KHERON::ROACHTANSTAAFL !Mon Sep 25 1989 15:19182
Printed by: Pat Roach                                    Document Number: 008797
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Sep-1989 11:09pm CET
                                        From:     DOUGLASS
                                                  DOUGLASS@AITG@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AI PARTICIPATION IN "GRAD SCHOOL" - PLEASE LET ME KNOW IF YOU WILL BE ABLE TO ATTEND  SO I CAN PLAN ON  HOW MANY WORKSTATIONS I'LL NEED...THANKS.  JEANNE

********************************************************************************

		AI PARTICIPATION IN DU:IT GRADUATE SCHOOL

********************************************************************************

Remember last year's AI SYMPOSIUM???  This year the US Sales and Sales Support
Organization has renamed the symposia as Digital UNIVERSITY: Institute 
of Technology. It is the logical extension of this summer's Digital 
University: Summer Program for Sales.  It will be a continuous training effort
for all field people throughout this fiscal year.

Our AI content is included under the CASE umbrella.  We will have an opportunity
not only to continue the education process for our own AI field consultants, but
to start the education process of a much larger field audience. Expected 
attendance is 200-250.  

As of last week we have confirmed times and slots approved. (The process for
electing sessions was similar to the DECworld process.)

	WHEN:	OCTOBER 29 - NOVEMBER 3
		
		AI SPECIFIC DAY - NOVEMBER 1
	
	WHERE:	CHELMSFORD DIGITAL FACILITY


Ken Olsen will open up each session of the Graduate School.   There will
be 4 separate sessions this fall -

	o Desktop
	o Realtime
	o Transaction Proc/ Database
      *	o CASE (/AI)  *****THIS IS OUR SYMPOSIUM******

At the beginning, each student will be tested on their individual level 
of knowledge about the product set and product environments.  During the 
week, as in Summer School, the student can go back and try to improve upon 
his/her individual score given the new information learned.

Therefore, the OBJECTIVE of DU:IT is to give product knowledge to sales support
to better prepare them to support a the overall sale.  The environment into 
which we sell our product is also of prime importance to equip sales support 
with the ability to recognize opportunities and to match situation with
product recommendations.

Here is the our session map for the AI Track on Nov. 1st.  These sessions
are confirmed so we now have the ability to fine tune with the input from
our AI field partners, product managers and our consulting partners. 



WEDNESDAY, NOV. 1ST

------------------------------------------------------------------------------
  TIME			SESSION	 				AUDIENCE	
-------------------------------------------------------------------------------	
8:00 - 10:00	    EXPERT SYSTEMS			 200-250 Sales Support
							 Specialists - Software
		The expectation is that the student	 Backgrounds both CASE
	        will be able to leave understanding      and AI
		the synergy between the AI and CASE 
		environments.
		
		Work the concept of "AI Enhanced" applications
		and integration.  Key points:

		o Using AI as added value in solving some
		  specific problems - diagnostics, configuration.
		o Looking at different SW methodologies/how AI fits what
		  they know.
		o Recognizing an opportunity to
		  propose an expert system solution.
		o Emphasizing the importance of selling
		  services.
		o Understanding what Digital sells. 
		o Understanding how it can be used as a
		  competitive advantage. 
		o DEVELOPING ENTHUSIASM

		PROPOSED SPEAKER: Jack Rahaim & Tech. Person
_______________________________________________________________________________ 
		

10:00 - 12:00	    EXPERT SYSTEM PRODUCT UPDATE	 20-30 AI CONSULTANTS
   			AND LAB SESSION			 AND CASE CONSULTANTS
     &
		The objective of this session is
1:00  - 2:00	for a technical review/update
		of our existing product set (overhead
		format) and a look at some key demos 
		(from IJCAI) with the intent of allowing 
		the sales support people to get some hands on
		experience.

NOTE:		EACH SALES SUPPORT PERSON WILL HAVE A VAXSTATION 3100
		ON THE DESKTOP WHICH WE PLAN TO LOAD WITH THE APPROPRIATE
		SOFTWARE.  DEADLINE FOR LOADING ALL SOFTWARE: OCTOBER 16.

		Existing product updates would include:
		o VAX LISP 
		o VAX OPS5, plus PLATFORM BANKING demo
		o VAX Decision Expert, plus 2 application demos
		
		3rd Party product updates would include JUST
		an overhead presentation..describing our direction
		and some product positioning.  NO THIRD PARTY
		REPRESENTATIVES WILL BE ASKED TO JOIN THIS SESSION.


*TUES. OCT. 31	(Third Party representatives will be asked to attend
5:15 PM 	 the BIRDS OF A FEATHER SESSION ON TUESDAY)
	

PROPOSED PRESENTERS:	LIZ SOOHO, MARCIA ROLAND, SHEELA PATEL, ENGINEERING
			REPRESENTATIVES, THEMIS PAPAGEORGE	

-------------------------------------------------------------------------

2:30 - 4:15	  EPITOOL PRODUCT TRAINING		20-30 AI AND CASE
							CONSULTANTS
		The objective of this session is	
		to introduce this product to our
		sales support people.  The first
		segment will address the company,
		product, and successes.  Second
		hour will be lab/classroom time.
		It is intended to give enough product
		familiarity to prepare sales support
		to discuss the product with a customer.		
		

PROPOSED SPEAKERS:  PERTTI JAUHIAINEN (EPITEC), MALIN GOODWIN
-------------------------------------------------------------------------------

4:15 - 5:15	  PRODUCTIVITY ENHANCEMENT SERVICES	     20-30 AI AND CASE
							     CONSULTANTS  
		  	
		The objective of this session is to 
		familiarize the sales support audience
		with the new Productivity Enhancement Services
		we introduced at our International Joint
		AI Conference this August.  We plan to demonstrate
		the internal tools being used to deliver this
		new service.

		There will also be an opportunity to discuss other
		Consulting Services issues, and exchange ideas and
		successes.


PROPOSED SPEAKERS:	DON WEYMOUTH, ED KILLEEN
________________________________________________________________________________

Keep in mind, this is the ONLY opportunity we get to train our AI field
consultants.  I would appreciate your cooperation and want to work with
you to do the best job possible for our dedicated...and much appreciated
field people.

I will pull together a core group of people over the next week to review 
and fine tune agenda items.  

Thanks. 

Jeanne
 
67.28KNOWLEDGE CRAFT TRAININGHERON::ROACHTANSTAAFL !Mon Sep 25 1989 15:3776
Printed by: Pat Roach                                    Document Number: 008808
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Sep-1989 10:38pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: KNOWLEDGE CRAFT TRAINING


		        ADVANCED AI TECHNICAL COURSE
  ***************************************************************************
			  Knowledge Craft Training
  ***************************************************************************

     DATE:     October 23 - November 3, 1989

  LOCATION:    Lab 3, DLB12 - 295 Donald Lynch Boulevard, Marlborough, MA

  DESCRIPTION: Knowledge Craft training is a two-week introduction to the
               Knowledge Craft product.  The course focuses primarily on the 
	       functionality of Knowledge Craft's components. Topics covered 
               during the training include:

	     o CRL Topics: Knowledge Craft shell and environment, schema and 
	       schema network editors, contexts, relations, switches,
     	       restrictions, user-defined relations, transitivity, 
	       inheritance, path restrictions, demons, meta knowledge,
	       dependency maintenance, error handling and the save
	       manager;

	     o Inferencing strategies: object programming, the agenda mechanism,
	       CRL-OPS and CRL-Prolog;

	     o CRL-OPS: forward chaining, rule-based programming style, conflict
	       resolution, CRL-OPS workcenter and CRL/CRL-OPS integration.

        GOALS: Upon completion of this course, participants will understand
               Knowledge Craft syntax including the:

	     o Knowledge Craft representational and inferencing strategies:
	       CRL (TM), CRL-OPS (TM), CRL-Prolog (TM), Object Programming
	       and Agenda Mechanism;

	     o Knowledge Craft workcenters: CRL workcenter, CRL-OPS workcenter
	       and the CRL-Prolog workcenter; as well as the environments,
	       editors, and utilities; and

	     o Interface Development Package: window/graphics system, command
	       system, and interface schema filler.

    AUDIENCE &
PREREQUISITIES: Programming experience with Lisp.  This course is designed 
		primarily for engineers who will use Knowledge Craft to design 
		and develop knowledge-based systems.  Attendees should have 
		prior experience with the Common Lisp language and the MicroVax 
		workstation.  Participants should also have a basic under-
		standing of knowledge-based systems.

          COST: $3600

 LENGTH&FORMAT: 2 weeks, Lecture/Lab

   TO REGISTER: Contact Louise King on AIADM::KING with your 1) name, 2)
		node, 3) organization, 4) manager's name and 5) cost center.
		
 *************************************************************************
            Class starts at 9:00 am on Monday, October 23rd.
 *************************************************************************
67.29HILL, 9/22, DLB5CAFE9:30HERON::ROACHTANSTAAFL !Mon Sep 25 1989 15:4671
Printed by: Pat Roach                                    Document Number: 008809
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Sep-1989 10:51pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: HILL, 9/22, DLB5CAFE9:30







	TITLE:  	"The Mind at AI:  Horseless Carriage to Clock"

	SPEAKER:  	DR. WILLIAM C. HILL
	  		Microelectronics and Computer Technology 
			Corporation (MCC)

	DATE:		Friday, September 22, 1989

	TIME:		9:30-11:00AM

	PLACE:		DLB5 Cafeteria

	HOST:  		BILL KUHLMAN
       			MCC liaison, Corporate Research

	This talk articulates how the intellectual activity we label
	"AI" can be likened, in revealing ways, to each of five familiar
	technologies.

	Commentators on AI converge on two goals they believe define 
	the field: (1) to better understand the mind by specifying 
	computational models, and (2) to construct computer systems 
	that perform actions traditionally regarded as mental.  

	We should recognize that AI has a third, hidden, more basic 
	aim; that the first two goals are special cases of the third; 
	and that the actual technical substance of AI concerns only 
	this more basic aim.  
	
	This third aim is to establish new computation-based represen-
	tational media, media in which human intellect can come to 
	express itself with different clarity and force.  This talk 
	articulates this proposal.

	Dr. Hill is a scientist in the MCC Human Interface Laboratory 
	in Austin, Texas.  He conducts research on mediating collabora-
	tion computationally, using knowledge-intensive interface 
	techniques.  

	He is part of the team evolving the Human Interface Tool Suite.
	This is a prototype tool environment for designing collaborative 
	interfaces to high-functionality environments.  It does this by 
	incorporating visual programming, restricted natural language, 
	gesture recognition, and task reasoning in desk-sized, multimodal 
	worksurfaces.

	William C. Hill received his Ph.D. from Northwestern Univer. for 
	his research on advice seeking, giving, and following of graphic 
	interfaces.
67.30 nowledge Craft CourseHERON::ROACHTANSTAAFL !Mon Sep 25 1989 16:4281
Printed by: Pat Roach                                    Document Number: 008826
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Sep-1989 10:19pm CET
                                        From:     DOUGLASS
                                                  DOUGLASS@AITG@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Knowledge Craft Course

From:	AIADM::TSS "20-Sep-1989 1131" 20-SEP-1989 12:10:29.83
To:	TSS_Distribution
CC:	
Subj:	KNOWLEDGE CRAFT TRAINING


		        ADVANCED AI TECHNICAL COURSE
  ***************************************************************************
			  Knowledge Craft Training
  ***************************************************************************

     DATE:     October 23 - November 3, 1989

  LOCATION:    Lab 3, DLB12 - 295 Donald Lynch Boulevard, Marlborough, MA

  DESCRIPTION: Knowledge Craft training is a two-week introduction to the
               Knowledge Craft product.  The course focuses primarily on the 
	       functionality of Knowledge Craft's components. Topics covered 
               during the training include:

	     o CRL Topics: Knowledge Craft shell and environment, schema and 
	       schema network editors, contexts, relations, switches,
     	       restrictions, user-defined relations, transitivity, 
	       inheritance, path restrictions, demons, meta knowledge,
	       dependency maintenance, error handling and the save
	       manager;

	     o Inferencing strategies: object programming, the agenda mechanism,
	       CRL-OPS and CRL-Prolog;

	     o CRL-OPS: forward chaining, rule-based programming style, conflict
	       resolution, CRL-OPS workcenter and CRL/CRL-OPS integration.

        GOALS: Upon completion of this course, participants will understand
               Knowledge Craft syntax including the:

	     o Knowledge Craft representational and inferencing strategies:
	       CRL (TM), CRL-OPS (TM), CRL-Prolog (TM), Object Programming
	       and Agenda Mechanism;

	     o Knowledge Craft workcenters: CRL workcenter, CRL-OPS workcenter
	       and the CRL-Prolog workcenter; as well as the environments,
	       editors, and utilities; and

	     o Interface Development Package: window/graphics system, command
	       system, and interface schema filler.

    AUDIENCE &
PREREQUISITIES: Programming experience with Lisp.  This course is designed 
		primarily for engineers who will use Knowledge Craft to design 
		and develop knowledge-based systems.  Attendees should have 
		prior experience with the Common Lisp language and the MicroVax 
		workstation.  Participants should also have a basic under-
		standing of knowledge-based systems.

          COST: $3600

 LENGTH&FORMAT: 2 weeks, Lecture/Lab

   TO REGISTER: Contact Louise King on AIADM::KING with your 1) name, 2)
		node, 3) organization, 4) manager's name and 5) cost center.
		
 *************************************************************************
            Class starts at 9:00 am on Monday, October 23rd.
 *************************************************************************
67.31NEURAL NET OFFERINGHERON::ROACHTANSTAAFL !Mon Sep 25 1989 16:4770
Printed by: Pat Roach                                    Document Number: 008827
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Sep-1989 10:59pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEURAL NET OFFERING

--------------------------------------------------------------------------------
                  ETE      Satellite Network Courses Description
________________________________________________________________________________
Title             NEURAL NETWORKS: CAPABILITIES TODAY FOR THE FUTURE          
Instructor(s)     SCHWARTZ/WILLOW/ANDERSON      
Date              27-SEP-89 - 27-SEP-89
Time              12:00 -  3:00
Location          UNIVERSITY VIDEO HLO2-1/D07   
Course Number     SAT01-27
Course Fee        $  500
Registration      Log into COURSES software on SHARE.
                  Username and Password are both COURSES.
Cancellation      PLEASE CANCEL 1 WEEK PRIOR TO THE COURSE START DATE TO      
                  AVOID BEING CHARGED.                                        


Instructors:      Tom Schwartz, Bernard Willow, James Anderson
                  
                  *  This is a Live Satellite Broadcast *
                                                                                
Description:      The historical development and fundamental ideas of
                  neural networks are used in advancing signal processing
                  technology.  How to apply neural network technology is
                  used to obtain a strategic competitive advantage.
                  Examples of neural network 'computation' demonstrating
                  the strengths and limitations of neural networks.  How
                  neural networks can be used for adaptive control of
                  non-linear systems.  Neural network case studies from
                  both technical and business points of view.
                  Characteristics of successful neural network problem
                  selection, including adding neural network technology to
                  existing products to improve performance.
                  
Speaker:          Tom Schwartz is an internationally recognized neural
                  network expert. He is the founder of the Schwartz
                  Associates, a six year-old consulting firm specializing
                  in the commercialization of advanced computer
                  technologies.
                                                                                
                                                                                
                                                                                
                  Dr. Bernard Widrow is the president of the International
                  Neural Network Society. Widrow is a professor of
                  electrical engineering at Stanford University.
                                                                                
                                                                                
                                                                                
                  James A. Anderson serves as a director of the
                  International Neural Network Society and is co-editor of
                  the book, "Neural Computing: Foundations of Research."
 
--------------------------------------------------------------------------------
                  Engineering Training & Education,
                  Supporting Semiconductor Operations
67.32PAO, 11/1, 9:30HERON::ROACHTANSTAAFL !Mon Sep 25 1989 17:1395
Printed by: Pat Roach                                    Document Number: 008835
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Sep-1989 02:22am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PAO, 11/1, 9:30









	TITLE:		"Adaptive Learning of a Dynamic Process"

	SPEAKER:	DR. YOH-HAN PAO		
			Case Western Reserve University and
			AI Ware, Inc., Cleveland, OH

	DATE:		Wednesday, November 1, 1989

	TIME:		9:30-11:30AM

	PLACE:		Pacific Room, DLB12

	HOST:		MATT JAFFE
			Business Manager
			Machine Learning and AI Services Group

	=============================================================
	NOTE:  This is an TSS THEME, and is the first in a series
	       of lectures in the area of machine learning.
	=============================================================

	The response of dynamic systems can be understood if we can model
	the internal mechanism of the system.  However, sometimes systems
	are complex and "opaque" and we cannot build up a causal model.	
	Social, financial and industrial systems typically all have parts
	which are opaque to some extent.

	Despite the lack of causal models, we do build up knowledge of
	how to deal with such systems.  Our present discussion deals with
	the task of teaching machines that same expertise so that they
	can also monitor, predict and control such partially opaque complex
	systems, hopefully better than we can.

	Specifically, in this discussion, we show how a machine can 
	acquire a QUANTITATIVE model of such a complex process by 
	observing instances of its behavior, either as it evolves
	freely, or in response to disturbances.

	The essence of the approach is that of adaptive pattern recogni-
	tion as implemented with neural-net computing.  The description
	built up in this manner can be generalized and specialized.  The
	machine is, therefore, able to track the NATURE of the system
	adaptively, even if that evolves with time.

	A qualitative description of neural-net based supervised learning
	is sufficient to provide a basis for understanding the conceptual
	framework of this discussion.







	"Adaptive Learning of a Dynamic Process"
	Dr. Yoh-Han Pao
	November 1, 1989




	Dr. Pao received his Ph.D. from Penn. State Univ. in 1952.  He
	has also studied at London Univ. and Univ. of Edinburgh.

	He is a Prof. of Electrical Eng. and Computer Science, a George
	S. Dively Distinguished Prof. of Eng., Director for the Center 
	for Automation and Intelligent Systems Research and President
	of AI Ware, Inc.

	His current research interests include the areas of adaptive
	pattern recognition and AI, as well as signal and image processing. 
67.33GEEP, 9/26, 9:30HERON::ROACHTANSTAAFL !Mon Sep 25 1989 17:2072
Printed by: Pat Roach                                    Document Number: 008836
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Sep-1989 02:37am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: GEEP, 9/26, 9:30






	TITLE:		"Graduate Engineering Education
                        Program"  (GEEP)

	SPEAKER:	LEE PETERSON
			GEEP Program Consultant

	DATE:		Tuesday, September 26, 1989

	TIME:		9:30-11:30 AM

	PLACE:		Pacific Room, DLB12

	HOST:		JUDE PARTRIDGE
			AITC TSS Program Manager	

	The GEEP provides the opportunity for engineers to pursue
	technical graduate degrees in engineering and computer 
 	science, at universities on a full-time basis.

	Those engineers who apply and are accepted into the GEEP,
	receive full salary and fringe benefits, tuition, books
	and fees, and the cost of relocation while they are 
     	studying at their respective universities.

        If you have been thinking about continuing your technical
	education at the graduate level, or if you have an employee
	who is/should be thinking about a full-time university 
        program, plan to attend a GEEP seminar to learn about:
	
			- how to apply
			- the selection process
			- appropriate university programs
			- program history and current status
     
	The application deadline for the 1990-91 academic year is
	15 February 1990.  It's not too soon to get the information
	you need to apply.  

                   ________________________________

	 Lee Peterson is the program consultant to the Graduate  
        Engineering Education Program (GEEP).

        In that capacity, Lee works with applicants during all phases 
        of the application and selection processes of the Program.  
        She also continues to work with GEEP participants during 
        their university experience and reentry to Digital.

        Lee's industry experience includes nearly ten years of working 
        with engineers in the areas of training and development, career 
        counseling, management development and employee relations. 
67.34RANDY DAVIS, 10/4, 9:30HERON::ROACHTANSTAAFL !Fri Oct 06 1989 15:3791
Printed by: Pat Roach                                    Document Number: 008946
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     03-Oct-1989 07:30pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: RANDY DAVIS, 10/4, 9:30









	TITLE:		"Model-Based Reasoning:  Recent 
	        	Results And Future Plans"

	SPEAKER:	PROFESSOR RANDALL DAVIS
			Associate Director, MIT AI Lab
			
	DATE:		Wednesday, October 4, 1989

	TIME:		9:30-11:30 AM

	PLACE:		Pacific Room, DLB12

	HOST:		MEYER BILLMERS
			Consulting Software Engineer
			AI Applications Group, AI Technology Center


	Recent work at MIT in model-based reasoning has produced results in
	several areas, including:

	*  techniques that allow our diagnostic system to learn from its 
	   experience by generalizing solved problems, in a synthesis of 
	   model-based reasoning and explanation-based generalization;

	*  modeling principles that permit representing and reasoning about 
	   board-level circuits with complex behavior, illustrated using a 
	   board containing 30 chips, including two microprocessors;

	*  an approach to circuit design capable of creating novel device 
	   topologies by reasoning from an understanding of how the device 
	   works.  Unlike the traditional approach of doing routine design 
	   by selecting component designs from a library, this system is 
	   capable of innovative design;

	*  a system capable of designing ``good'' representations for a 
	   task from a knowledge base of representations and information 
	   about when each is appropriate


	Future work will focus on issues such as model selection, i.e., 
	determining how to ``think about'' a device.  How is it that we 
	know what properties of the device to consider and which to ignore?  
	We also speculate on what it would take to create an interactive, 
	automated version of a book like ``The Way Things Work,'' as a way 
	of understanding what we mean when we say we understand how something 
	works.







	"Model-Based Reasoning:  Recent
	Results and Future Plans"
	Professor Randall Davis
	Page Two



	Randall Davis is a Professor in both the School of Management and 
	the Electrical Engineering and Computer Science Department, as well 
	as Associate Director of the Artificial Intelligence Laboratory.  
	
	He and his group at MIT have developed model-based systems for 
	troubleshooting, generating diagnostics, design for testability, 
	innovative design, and other functions.
67.35KNOWLEDGE REPRESENTATIONHERON::ROACHTANSTAAFL !Fri Oct 06 1989 15:4072
Printed by: Pat Roach                                    Document Number: 008949
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     04-Oct-1989 05:10am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: KNOWLEDGE REPRESENTATION


********************************************************************************
		          ADVANCED COURSE OFFERING
********************************************************************************

 		      ADVANCED KNOWLEDGE REPRESENTATION

	       DATE:    Nov. 14, 17, 21, 28, Dec. 5, 12, 15, 19, 1989.
	                (2.5 hour per session from 1:00 - 3:30 p.m.)

	   LOCATION:    Discovery Conference Room - DLB12-1/AI
			Digital Equipment Corporation
			295 Donald Lynch Boulevard
			Marlborough, MA  01532

	DESCRIPTION:	This course will survey and discuss major techniques
			and issues in Knowledge Representation.  For each 
			representation, the subtleties, strengths and 
			weakness will be discussed.  For each meeting, every 
			participant will be expected to write a short critique
			of the week's readings which will be distributed to
			all participants.  In addition, at each meeting, a
			participant will give a short presentation with the 
			instructor, on the major important or controversial 
			highlights of the readings assigned.  Sources for the
			readings include: Readings in Artificial Intelligence,
			by Webber & Nelson; Reading in Knowledge Repre-
			sentation, Brachman & Levesque; and Handbook of AI, 
			by Bars & Fergenfaum (eds.).

	       GOALS: 	Upon completion of the course participants will:

			o Understand current Knowledge Representation research 
			  issues.

			o Be aware of alternative Knowledge Representation 
			  techniques, and their appropriateness in applications 
			  development.

			o Be deeply familiar with classical and influential 
			  research papers in the K.R. field.

	    PRESENTER:	Dr. Edwina Rissland, Professor, U.MASS, Amherst 
			Computer & Information Systems

      PRE-REQUISITES:  A good understanding of traditional Knowledge 
			Representation	techniques, and an ability to read, 
			discuss and present research papers at the graduate 
			level.

	         COST:  $1200 for internal pepole.
	     	      
	       FORMAT: 	Lectures/presentations/discussions/written critiques.

	  TO REGISTER:  Send mail to AIADM::KING.  Please include badge number,
			DTN, group name, organization name and manager's name.
67.36CAPPUCCILLI, 10/5, 9:30HERON::ROACHTANSTAAFL !Fri Oct 06 1989 15:4193
Printed by: Pat Roach                                    Document Number: 008950
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     04-Oct-1989 05:10am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: CAPPUCCILLI, 10/5, 9:30








        
	TITLE: 		"An Introduction to Artificial Intelligence
                        For the Non-Technical Person"

        SPEAKER:  	TERRY CAPPUCCILLI 
	                Information Specialist
                   	Artificial Intelligence Information Center
			AI Technology Center

	NOTE:		KAHU CHEN of the Applied Intelligent
                        Systems Group, AITC,  will be on hand 
			to assist with questions.
 
        DATE:  		Thursday, October 5, 1989

        TIME:  		9:30-11:00AM

        PLACE:		Pacific Room DLB12

	HOST:		Jude Partridge
			AITC TSS Program Manager

        Computers are invaluable in helping people to organize,
        reason and solve problems.  And Artificial Intelligence has
        much to contribute in extending the power of computers
        beyond their basic mathematical and statistical functions.
        The hype and myth that once surrounded AI are now beginning
        to disiipate and more realistic expectations and possibilities
        are within sight.

        Intended as an introduction for the non-technical person,
        this presentation will address a number of key concepts
        associated with AI - what it is, how it's used and what its
        impact will be.  Expert Systems, the most well known area of
        AI, will also be looked at since they are literally
        changing the way businesses operate and solve problems.
        What they are, how they function and examples of current
        applications will be discussed to heighten your awareness
        of these remarkable programs.

        And finally, Digital's involvement in the technology, which
        has resulted in real world applications, products and services
        will be outlined.  Digital has more experience with Expert
        systems than anyone else.  It also has learned more about
        how to develop and not to develop Expert Systems than most
        companies.







	"An Introduction to Artificial Intelligence
	For the Non-Technical Person"
	October 5, 1989
	Page Two




        Terry Cappuccilli is an Informatin Specialist at The
        Artificial Intelligence Information Center at DLB5 and
        has been employed by Digital for five years. She holds an
        MEd degree in Education from Boston College and an MS degree
        in Library and Information Science from Simmons College.


	NOTE:  The AI Information Center will make available
               an additional selection of AI-related articles.
67.37LIPKA, 10/12, 9:30AMHERON::ROACHTANSTAAFL !Thu Oct 12 1989 11:27119
Printed by: Pat Roach                                    Document Number: 009036
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Oct-1989 08:47pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: LIPKA, 10/12, 9:30AM







	TITLE: 		"Using Models for Engineering Systems"

	SPEAKER:	DR. STEPHEN LIPKA
		        Advanced Systems and Tools Group
			AI Technology Center

	DATE:		Thursday, October 12, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room, DLB12

	HOST:		CHRIS CHERPAS
			AI Training and Education Group
			AI Technology Center

	
	This talk will present a limited set of model types, including 
	IDEF0 (functional) modeling and Entity-Relationship-Attribute 
	(information) modeling. Their capabilities will be presented, 
	and their range of usefulness will be discussed. 

	An example drawn from a software prototype developed within 
	Digital will illustrate their use. A work-flow model developed 
	for that prototype will also be presented, and the encoding of 
	that model into a work-flow manager will be briefly discussed.
	

	Background:
	----------

	If one were to characterize the spectrum of single-person 
	software development efforts, one would likely find that the 
	developer was knowledgeable either about the domain being 
	supported or about software. 

	The best of both worlds occurs when a software-knowledgeable 
	developer is adequately knowledgeable about the domain being 
	supported.  This usually occurs when the developer creates a 
	model (perhaps mental) of the domain. 

	For larger projects to succeed, it is also necessary that all 
	members within any one group (domain-knowledgeable or software-
	knowledgeable) agree, i.e., have the same model in mind. 

	Formal models are a useful tool in developing software systems. 
	Useful models can describe (business or software) "systems" 
	rigorously, to some well understood tolerance. Such models are 
	useful for conveying understanding within and across teams. 







	"Using Models for Engineering Systems"
	Dr. Stephen Lipka, October 12, 1989
	Page Two



	
	Models provide visibility into the learning or design process 
	and provide a basis for organized discussions. Well-chosen 
	models make it easier to judge the quality of a "system" when 
	judging by the appropriate design rules. And some models may be 
	encoded in such a way as to drive execution of a system or allow 
	machine-based reasoning.

	Throughout the system development process, a wide spectrum of 
	models can be profitably applied. Some are more useful for 
	business.  Some are appropriate for detailed software.  But, 
	few are appropriate for a wide range of purposes. 

	The important issues in choosing a model type, therefore, are 
	understanding what problem to solve, understanding what a given 
	type of model can do, and understanding how a given model type 
	can be usefully applied. 
	
	Biography:
	---------

	Stephen Lipka is an independent consultant.  He is interested in 
	the use of static and dynamic models for software engineering, 
	multiple-model techniques, and CASE tools that integrate across 
	models. 

	He received a Bachelor's Degree in Physics and a Master of 
	Science in Electrical Engineering at the Polytechnic Institute 
	of Brooklyn. He received his Ph.D. in Computer Science from the 
	State University of New York at Stony Brook in 1975. 

	Stephen is currently working with the Applied Systems and Tools 
	Group at the AI Technology Center in Marlborough, MA. He has a 
	large scope of experience in applying and teaching modeling and 
	software engineering techniques. 
67.38BROWN BAG SEMINAR 10/12HERON::ROACHTANSTAAFL !Thu Oct 12 1989 11:2857
Printed by: Pat Roach                                    Document Number: 009037
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Oct-1989 01:27am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: BROWN BAG SEMINAR 10/12

	TITLE:		DISTRIBUTED AI and the MULTI-AGENT COMPUTING 
	                ENVIRONMENT (MACE)

	SPEAKER:	Professor Les Gasser

	DATE:		Thursday, 12 October, 1989

	TIME:		12 Noon to 1:00 PM  (Brown Bag Seminar)

	PLACE:		Pacific Room, DLB12

	HOST:		Chuck Marshall
			Advanced Systems and Tools Group
			AI Technology Center

Because of their ubiquity and promise, research interest in intelligent
multi-agent systems and in Distributed AI in general is growing explosively.
But there has been relatively little conceptual research on problems of
CONCERTED ACTION or COORDINATION in multi-agent systems.  In distributed
computing systems, for example, programmers are given tools, mechanisms, and
knowledge with which they create coordination protocols. Programs then ENACT
the protocols programmers have devised. In distributed intelligent systems,
however, the problem is larger: how can the agents in system themselves treat
mutual coordination as a set of problems to be recognized, stated, and solved? 

In this talk I will present a some results from our research into description
and implementation structures for multi-agent systems, including some
requirements for concerted action, structures for specifying and representing
agents and their concerted action, and approaches to forming and maintaining
alliances among self-interested agents.  We will give particular attention to
strategies used in the USC DAI testbed, MACE.  This work provides a basis for
understanding or designing interactions in many kinds of DAI systems, such as
cooperating expert systems, multi-agent planning or diagnostic systems,
person-machine cooperative systems, etc. 

Distributed Artificial Intelligence Group
Dept. of Computer Science, SAL-200
University of Southern California
Los Angeles, CA 90089-0782
(213) 743-7794
[email protected]
67.39RESENDHERON::ROACHTANSTAAFL !Fri Oct 13 1989 14:1871
Printed by: Pat Roach                                    Document Number: 009055
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Oct-1989 01:46am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: RESEND

********************************************************************************
		          ADVANCED COURSE OFFERING
********************************************************************************

 		      ADVANCED KNOWLEDGE REPRESENTATION

	       DATE:    Nov. 14, 17, 21, 28, Dec. 5, 12, 15, 19, 1989.
	                (2.5 hour per session from 1:00 - 3:30 p.m.)

	   LOCATION:    Discovery Conference Room - DLB12-1/AI
			Digital Equipment Corporation
			295 Donald Lynch Boulevard
			Marlborough, MA  01532

	DESCRIPTION:	This course will survey and discuss major techniques
			and issues in Knowledge Representation.  For each 
			representation, the subtleties, strengths and 
			weakness will be discussed.  For each meeting, every 
			participant will be expected to write a short critique
			of the week's readings which will be distributed to
			all participants.  In addition, at each meeting, a
			participant will give a short presentation with the 
			instructor, on the major important or controversial 
			highlights of the readings assigned.  Sources for the
			readings include: Readings in Artificial Intelligence,
			by Webber & Nelson; Reading in Knowledge Repre-
			sentation, Brachman & Levesque; and Handbook of AI, 
			by Bars & Fergenfaum (eds.).

	       GOALS: 	Upon completion of the course participants will:

			o Understand current Knowledge Representation research 
			  issues.

			o Be aware of alternative Knowledge Representation 
			  techniques, and their appropriateness in applications 
			  development.

			o Be deeply familiar with classical and influential 
			  research papers in the K.R. field.

	    PRESENTER:	Dr. Edwina Rissland, Professor, U.MASS, Amherst 
			Computer & Information Systems

      PRE-REQUISITES:  A good understanding of traditional Knowledge 
			Representation	techniques, and an ability to read, 
			discuss and present research papers at the graduate 
			level.

	         COST:  $1200 for internal pepole.
	     	      
	       FORMAT: 	Lectures/presentations/discussions/written critiques.

	  TO REGISTER:  Send mail to AIADM::KING.  Please include badge number,
			DTN, group name, organization name and manager's name.
67.40EPITOOL QUESTIONS?HERON::ROACHTANSTAAFL !Fri Oct 13 1989 21:1152
Printed by: Pat Roach                                    Document Number: 009072
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Oct-1989 08:07pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL QUESTIONS?

               _______________________________________________

         	D O    Y O U    H A V E    Q U E S T I O N S	

                	A B O U T    E P I T O O L ?

               _______________________________________________

         	Meet Kristina Ernstsson and Bj�rn M�ller from
         	Epitec Sweden who are visiting Digital October 16 - 
         	October 20.   

         	Bj�rn has several years of experience in knowledge
         	engineering and applications development using
         	Epitool.

         	Kristina is responsible for Epitool training at 
         	Epitec AB.


         	Location:	Discovery Conference Room
         			DLB12, second floor, Pole  D2
         			Marlboro

         	Time:		Monday October 16
         			2 PM - 4 PM
		
		____________________________________________________
 
        			W E L C O M E ! ! !	
	       	____________________________________________________ 
 
         	Malin Goodwin, DTN 291-8894
         	Educational Services Intelligent Systems

         	 
67.41A NEW CHARTER - Presenting the AITC Presentation SeHERON::ROACHTANSTAAFL !Wed Oct 18 1989 11:06386
Printed by: Pat Roach                                    Document Number: 009095
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Oct-1989 01:30am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: A NEW CHARTER - Presenting the AITC Presentation Series

   +---+---+---+---+---+---+---+ TM 
   | d | I | g | I | t | a | l |       INTEROFFICE MEMORANDUM		Page 1
   +---+---+---+---+---+---+---+ 

   TO: AITC Personnel                  FROM:  JUDE PARTRIDGE
                                       DATE:  17 October, 1989
                                       DEPT:  AITC Presentation
                                              Series Program Manager
                                       EXT:   291-8252
                                       LOC:   DLB12-2/D3
                                       ENET:  AIADM::TSS 
   ====================================================================
   SUBJECT:   AITC's PRESENTATION SERIES
   ====================================================================

   You are cordially invited to get to know the AITC Seminar Series
   better!

   Our people are AITC's most important resource. The Seminar Series is
   a mechanism to leverage the expertise and productivity within the AI 
   Technology Center.  It provides an opportunity to maximize our resources
   by means of the following:

		1.  Technical Seminar Series
		2.  AITC Seminar Series
		3.  Development Workshops/Courses (AITC-Specific)
		4.  Brown Bag Seminars

   Benefits to AITC:
   ----------------
		o Information flow among groups (cross-AITC 
	          communication);
			
		o Increase overall technical, business and
                  personal competencies throughout the TC;
		
		o Identifying AITC technology opportunities 
                  for collaboration; 
		  
		o Stay focused on the technology as outlined 
	          by AITC's senior technical management
     
   Benefits To Presenters:
   ----------------------
		o Visibility to individual projects;
		
		o Display value to the AITC;
			
		o Personal growth and development:
			
			o Hone presentation skills
			
			o Performance review input
				
			o Open doors to the future			








			     TABLE OF CONTENTS
                        



								Page
                                                        	----



		1.  Technical Seminar Series . . . . . . . . . .  2

		2.  AITC Seminar Series. . . . . . . . . . . . .  3

		3.  Development Workshop/Courses (AITC-Specific)  4

		4.  Brown Bag Seminars . . . . . . . . . . . . .  5

		Examples by Category . . . . . . . . . . . . . .  6
		




	AITC PRESENTATION SERIES                                  	Page 2


		1.	TECHNICAL SEMINAR SERIES 
			------------------------
	Definition:
	----------
		o a means for technical information exchange and transfer 
	          to the AI engineering community at Digital
	
	Presenters:
	----------
	        o internal to Digital
		o external from industry or academia
	
	Hosts:
	-----
		o engineers with technical expertise

	Topics:
	------
		o supports the goal of the AITC to promote
                  AI throughout the Corporation
		o interest and impact to world-wide audience
	
	Notice Required:
	---------------
		o maximum possible - 3-4 weeks-->several months
		o minimum okay, depending on availability of
	          speaker and videotapes.

	Publicity:
	---------
		o worldwide electronic mail distribution
		o hard copy posted
		o "Digital This Week" (month's lead-time)
		o LIVEWIRE (month's lead-time)
==========================================================================
	o  Usually one to two hours in length, with an occasional half or 
	   whole day.  

	o  Location is usually Pacific Room, DLB12, or DLB5 Cafe.

	o  Often, additional technical meetings are planned for external 
	   speakers to provide greater opportunities for information 
	   exchange.

	o  Most seminars are videotaped, unless otherwise noted.

	o  Videotapes are made available to the Digital Library Network
	   (DLN), through the Hudson Library (SHARE::PANDERS), and the
           AI Information Center (AIADM::AIIC) at DLB5.

	o  Copies of handouts are usually sent to requesters after the 
	   seminar by the Program Manager.

	o  Most arrangements for seminar and speakers are made by the
	   Program Manager, Jude Partridge (AIADM::TSS).



	AITC PRESENTATION SERIES					Page 3


		2.	   AITC SEMINAR SERIES
			   -------------------

	Definition:
	----------
		o A structure for cross-AITC communication.  It is
                  designed to address topics relevant to the technical
	          and business goals, as outlined by the AITC's senior
	          staff.	
				
	Presenters:
	----------
	        o AITC business and technical people or associates; i.e.,
                  business partners, customers, etc.
			
	Topics:
	------
		o technical, business or personnel information applicable 
	          to AITC work
	
	Notice Required:
	---------------
		o two weeks preferred, depending on conference room 
	          availability

	Publicity:
	---------
		o AITC distribution lists, by Program Manager (ONLY)
		  (AIADM::TSS).

==========================================================================

	o  Seminar length dependent on topic; possible candidate for 
	   brown bag seminar.

	o  May take place in any available conference room at DLB5/12.
	   (Audience size usually 15-25 people.)

	o  Videotaped only as noted.

	o  If videotaped, tapes available via the AI Information Center 
	   (AIADM::AIIC) at DLB5; no extra copies made.

	o  Copies of handouts are available only as noted.

	o  Arrangements for seminar and speakers usually done as cooperative 
	   effort, working with Program Manager. It will be the responsibility 
	   of the sponsor to work out logistics of delivery in advance with 
	   the Program Manager, Jude Partridge (AIADM::TSS).



	AITC PRESENTATION SERIES					Page 4


		3.       DEVELOPMENT - WORKSHOPS/COURSES (AITC-Specific)
                         -------------------------------

	Definition:
	----------
		o AITC-customization of regular corporate offerings
                  or information-sharing offerings.

	Presenters:
	----------
		o internal to Digital
		o external from industry or academia

	Topics:
	------
		o technical - AITC specific
		o non-technical - AITC contained

	Notice Required:
	---------------
		o two weeks to several months in advance

	Publicity:
	---------
		o AITC distribution lists, accompanied by "AITC Only"
	          notice
============================================================================
	o Length of workshop/course dependent on topic - 1/2 to 2-3 days.

	o Registration is required in advance. Registrar may differ 
	  for each offering and will be noted on announcement.

	o Invitation limited to AITC cost centers.  Fee may be charged, and
          will be noted on announcement.  If offering is not filled, it will 
	  be opened to outside AITC cost centers.  Outside AITC cost centers 
	  will always be charged fee for attendance, even if offered at no
          charge to AITC personnel.

	o Location may be Pacific Room, DLB12, or any other available 
	  conference room at DLB5 or DLB12.

	o Videotaped only as noted.

	o If videotaped, tapes available via the AI Information Center at
	  (AIADM::AIIC) DLB5; no extra copies made.

	o Copies of handouts available only as noted.

	o Most arrangements for seminar and speakers are made by the Program 
	  Manager; however, cooperative efforts are welcomed. It will be the 
	  responsibility of the sponsor to work out logistics of delivery in 
	  advance with the Program Manager, Jude Partridge (AIADM::TSS).   



	AITC PRESENTATION SERIES					Page 5


		4.	  BROWN BAG SEMINARS
                          ------------------

	Definition:
	----------
		o more informal round table discussions to generate
	          cross-functional sharing and collaboration;
		o usually held at lunch time 

	Presenters:
	----------
		o internal to Digital
		o external from industry or academia

	Topics:
	------
		o technical:
			o AITC specific audience
				o AITC distribution lists
			o interest and impact to world-wide audience
				o worldwide electronic mail distribution

		o non-technical issues:
			o business
			o personnel
				o AITC specific 
					o AITC distribution lists
				o general interest/impact
					o worldwide electronic mail dist.

	Notice Required:
	---------------

		o 1 week preferred; however, sometimes this is the only
                  way to do last minute information sharing
==========================================================================
	o  Seminar length usually 1 hour.

	o  May take place in Leopard Room, DLB5, or any available conference 
	   room at DLB5/12. (Audience size usually 15-25.)
		
	o  Videotaped as noted.

	o  If videotaped, tapes available via the AI Information Center 
	   (AIADM::AIIC) at DLB5; no extra copies made

	o  Copies of handouts are available only as noted.

	o  Arrangements for seminar and speakers usually done as cooperative 
	   effort, working with Program Manager. It will be the responsibility 
	   of the sponsor to work out logistics of delivery in advance with 
	   the Program Manager, Jude Partridge (AIADM::TSS).




	AITC PRESENTATION SERIES					Page 6


			EXAMPLES OF OFFERINGS BY CATEGORIES
			-----------------------------------



	1.  Technical Seminar Series:
            ------------------------

		a.  Theme of 4 seminars on Object-Oriented Programming,
                    3/88
		
		b.  "Natural Language Research at the MCC Human 
                    Interface Lab", Dr. Kent Wittenburg, 7/22/88

		c.  "Adaptive Pattern Recognition (Machines
                    That Learn)", Ruby Li, 10/25/88

	2.  AITC Seminar Series:
	    -------------------

		a.  "Standards Awareness", Eleanor Feltes, 8/9/88

		b.  "New CGI/Digital Co-Developed Modules for 
	            Knowledge Craft", Cindy Cosic, 7/19/89

		c.  "Introduction to Neural Nets", Kahu Chen and
                    Professor Oded Maimon, 7/25/89

	3.  Training and Development:
	    ------------------------

		a.  "Communications at AITC", Chris Cole, 8/23/89	
		
		b.  "C Programming", 9/89

		c.  "The Balancing Act", Dr. Gayle Carson, 10/25/89

	4.  Brown Bag Seminars:
	    ------------------

		a.  "Assertive Communication", Marsha Greenberg, 7/27/89

		b.  "Poor Man's Hypertext - Integration VMSHelp Into
                    Your Application", John Comella, 6/14/89

		c.  "Technology Transfer - The MOC Experience", Mary
                    Lewis, 7/31/89
67.42DR. FREY, 11/15, 1PMHERON::ROACHTANSTAAFL !Wed Oct 18 1989 11:0694
Printed by: Pat Roach                                    Document Number: 009096
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Oct-1989 03:16am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. FREY, 11/15, 1PM







	TITLE:		"Exemplar-Based Decision Processes 
			in Machine Learning"

	SPEAKER:	DR. PETER W. FREY
			Northwestern Univ. and Odesta Corp.

	DATE:		Wednesday, November 15, 1989

	TIME:		1:00-3:00 PM

	PLACE:		Pacific Room, DLB12

	HOST:		MATT JAFFE
			Business Manager
			Machine Learning and AI Services Group


	===========================================================
	NOTE:  This is a TSS THEME, and is the second in a series
               of lectures in the area of machine learning.
	===========================================================


	Current expert systems summarize knowledge about a specialized
	domain by creating a list of IF-THEN production rules, and 
	then attempt to solve specific problems by using these rules 
	to reason in a deductive fashion from a set of givens.

	Human experts, in contrast, appear to store many years of 
	relevant experiences in relatively undigested form, and solve
	specific problems by reasoning in an analogical fashion, by
	noting similarities between the current problem and past
	experiences.
	
	Dr. Frey's work with exemplar-based decision systems attempts
	to model human problem solving strategies within the framework
	of conventional machine learning techniques.  In particular,
	he and his colleagues have been applying their own variations
	of the adaptive classifier, binary decision tree and nearest
	neighbor methodologies to a number of commercial applications.

	This talk will address the range of applications to which these
	methods are applicable, standard methods for representing the
	problems, decscriptions of their variations on the standard 
	methodologies, advantages and disadvantages of these approaches, 
	and the results of "head-to-head" comparisons of the methods 
	on a difficult character recognition problem.







	11/15/89
	Dr. Frey
	Page Two




	Peter W. Frey received his BA, magna cum laude, in 1964 from 
	Yale Univ., and his MA in 1966, Ph.D. in 1968 from Univ. of
	Wisconsin.

	He is a Prof. of Psychology at Northwestern Univ. and a senior
	research scientist at the Odesta Corp.

	Current research activities include, (1) expert systems which 
	learn from experience: applying a genetic algorithm to bit-mapped 
	production systems, and (2) exemplar-based decision processes:
	nearest neighbor and binary classification tree approaches.
67.43NOVEMBER CALENDARHERON::ROACHTANSTAAFL !Wed Oct 18 1989 11:0951
Printed by: Pat Roach                                    Document Number: 009097
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Oct-1989 04:48am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NOVEMBER CALENDAR






			

		AI TECHNOLOGY CENTER PRESENTATION SEMINAR SERIES

	
				    NOVEMBER
				    --------


	DATE/TIME/PLACE		SPEAKER/AFFILIATION	TITLE
	---------------         -------------------	-----

	11/1, Wednesday		Dr. Yoh-Han Pao		"Adaptive Learning 
	9:30-11:30AM		Case Western Reserve	of a Dynamic
	DLB12, Pacific Room	Univ. + AI Ware, Inc.   Process"

	11/3, Friday		Rager Ossel		"OSL MAPPER - 
	9:30-11:30AM		Open Systems Liaison	Integrating Mfg.
	DLB12, Pacific Room	Integrators BV		Cells with Expert
							Systems"

	11/15, Wednesday	Dr. Peter W. Frey	"Exemplar-Based
	1:00-3:00PM		Northwestern Univ. 	Decision Processes
	DLB12, Pacific Room	and Odesta Corp.	in Machine Learning"

	11/29, Wednesday	Dr. William Clancey	"Model Construction
	9:30-11:00AM		Institute for Research  Operators (Represent-
	DLB12, Pacific Room	on Learning		ing Control Knowledge
							for Generic Expert
							System Shells)"
67.44Announcing an advanced course offering. InterestedHERON::ROACHTANSTAAFL !Tue Nov 21 1989 09:2699
Printed by: Pat Roach                                    Document Number: 009137
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Nov-1989 08:34am CET
                                        From:     ROACH
                                                  ROACH@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Announcing an advanced course offering.  Interested?

 ****************************************************************************   
            BUILDING LARGE SCALE TIME CRITICAL EXPERT SYSTEMS
*****************************************************************************

       DATE:   December 13th and 14th, 1989

   LOCATION:   "Pacific" Conference Room - DLB12-2

DESCRIPTION:   The Program is designed to give you a thorough 
               understanding of how to build real time expert 
               systems, this new seminar covers everything from 
               concepts and technology issues, to the tools, 
               techniques, and methodologies involved.  It is 
               especially timely in light of the following 
               findings.   Today, industry's main artificial 
               intelligence focus is on expert systems - systems 
               that commonly adhere to the philosophy of 
               incremental development - and the single big 
               expert system model used in the world.  However, 
               this model has been found sub-optimal for 
               addressing large scale real time problems, both 
               architecturally and as a design example.  Further, 
               the issues involved in building expert systems are 
               different from those associated with building 
               small knowledge systems which have loosely 
               coupled characteristics to an overall environment 
               and application.  This intensive two-day seminar 
               will focus on these issues and will include 
               discussions that will be particularly useful to 
               engineers, developers, and managers alike:
          
               *  Artificial intelligence technology issues 
                  and operational issues are explored in detail, 
                  and lead to a sound useful methodology. 

               *  Case studies delve into such practical 
                  uses as alarm management, scheduling, stock 
                  trading, and statistical quality control.
     
               *  A set of examples illustrates the important 
                  concepts from the Process Control, CIM, 
                  financial, MIS/DP, and government systems 
                  areas.
     
               *  The MACRO artificial intelligence functions of 
                  control, simulation, estimation, and training 
                  are examined.

      GOALS:   Upon completion of the course, participants will 
               have an in depth understanding of:

               *    AI Technology Issues
                    Cooperating expert systems
                    Distributed AI
                    Domain dependent solution shells
                    Time critical AI

               *    AI Operational and Other Issues
                     An organizational structure
                     Hardware environments
                     Software environments and tools
                     Domain and problem selection
                     Systematic approach and checklist
                     AI operating system to the application
                      Reliability
                      Performance
                      Verification and validation
                      Conflict resolution
     
               *    Macro AI Functions
                     Control
                     Simulation
                     Estimation
                     Training

 PRESENTER:   Mike Stock

      COST:   $1,000.

LENGTH & FORMAT:  2 days, Lecture
 
TO REGISTER:  Send badge number, cost center, group and organizational
              name to AIADM::LEVIN.
67.45king - Please come,HERON::ROACHTANSTAAFL !Tue Nov 21 1989 09:3858
Printed by: Pat Roach                                    Document Number: 009143
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Nov-1989 08:50am CET
                                        From:     ROACH
                                                  ROACH@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: king - Please come,

********************************************************************************
	            DESIGNING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

OFFERINGS:	January 11, 1990
		March 21, 1990
		May 30, 1990

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:  	Knowledge engineers need to produce specifications for the 
		effective development of knowledge-based systems (KBSs).  
		These may be in the form of running prototypes and/or an 
		evolving set of design documents.  The purpose of this course 
		is to provide guidelines for balancing these two alternatives 
		and ensuring that critical design decisions are made regardless 
		of the method used.  The course will focus on the creation of 
		design documents which facilitate the optimal use of proto-
		typing as part of the design process.

GOALS:  	Upon completion of the course, participants will be able to:

		o Recognize the value of design for guiding KBS 
		  development
		o Describe a process which incorporates both planning 
	          and prototyping
		o Identify the most critical design decisions that affect 
	          KBS development
		o Contribute to the creation of a KBS design document

INSTRUCTOR:	Christopher Cherpas

LENGTH: 	1 day

FORMAT: 	Lecture/Discussion

COST:   	$500 

	
67.46DR. CLANCEY, 11/29, 9:30AMHERON::ROACHTANSTAAFL !Wed Nov 22 1989 15:53107
Printed by: Pat Roach                                    Document Number: 009175
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Nov-1989 07:26am CET
                                        From:     ROACH
                                                  ROACH@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. CLANCEY, 11/29, 9:30AM







	TITLE:		"Why Every Expert System Needs a Blackboard"

	SPEAKER:	DR. WILLIAM J. CLANCEY
			Senior Research Scientist
			Institute for Research on Learning (IRL)

	DATE:		Wednesday, November 29, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room DLB12

	HOST:		DR. JOHN MCDERMOTT
			AI Research Group Manager
			AI Technology Center

	AI programming is a technique for modelling processes qualitatively.
	
	A program's solution can be viewed as a model of some system in the
	world that the program is diagnosing, configuring, controlling, etc.
	
	The program's inference procedure can be viewed as operators that
	construct a graph-representation of a situation-specific model.
	
	Thus, we discover that the terminology of blackboard expert systems
	is not specific to a particular set of programs, but is rather a 
	valuable perspective for understanding what every expert system is
	doing.

	Differnt perspectives of expert system reasoning will be reviewed.
	A formal framework for describing Neomycin's model construction
	operators will be provided.
	
	Examples from TEIRESIAS, ABEL, Caduceus, HASP, ACCORD and EXPLAIN
	illustrate the unifying and clarifying nature of the model-con-
	struction perspective.  Further examples from KNACK and TOPO
	illustrate how to write metarules without using domain-specific 
	terms, thus making explicit their model-construction nature.

	Finally, a grand synthesis that unifies representation and control
	of processes in the three levels of domain, inference, and commun-
	ication modelling found in AI programs will be provided.

	Dr. William J. Clancey received his Ph.D. in Computer Science 
	from Stanford University, where he was later a Senior Research
	Associate at the Knowledge Systems Laboratory (1979-1987).







	Dr. William J. Clancey
	November 29, 1989
	Page Two




	He has been active in expert systems research since he joined
	the MYCIN project in 1975, for which he was co-developer of
	the antibiotic therapy and question-answering programs.

	His interests lie in computational modelling of cognition and
	the design of architectures for expert systems to facilitate 
	their construction, explanation and multiple use.  To investigate
	these issues, he has developed NEOMYCIN and instructional programs
	based upon it.

	Dr. Clancey has published widely on expert systems methodology.
	He is the author of "Knowledge-Based Tutoring:  The GUIDON
	Program" (MIT Press, 1987) and the co-author (with E. H. 
	Shortliffe) of "Readings in Medical Artificial Intelligence:
	The First Decade" (Addison-Wesley, 1984).  

	His paper on Heuristic Classification was nominated for the
	publisher's prize at AAAI-84.  He has presented several
	tutorials at AAAI and IJCAI on expert systems and educational
	applications and was Tutorial Chair for AAAI-87.  

	He is a member of the editorial board of several journals,
	including Cognitive Science and Artificial Intelligence. He is
	a councilor of the American Association of Artificial Intelli-
	gence, and Editor-in-Chief (exofficio) of the AAAI Press.		
67.47DR. LENAT, 12/5, 9:30AMHERON::ROACHTANSTAAFL !Wed Nov 22 1989 15:54152
Printed by: Pat Roach                                    Document Number: 009176
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Nov-1989 07:26am CET
                                        From:     ROACH
                                                  ROACH@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. LENAT, 12/5, 9:30AM





	TITLE:		"AI Versus Common Sense"

	SPEAKER:	DR. DOUGLAS B. LENAT
			Principal Scientist and Director
			MCC AI Laboratory

	DATE:		Tuesday, December 5, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room, DLB12

	HOSTS:		DR. JOHN MCDERMOTT, Manager, AI Research Group
			AI Technology Center, and
			BILL KUHLMAN, MCC Corporate Research Liaison


	Note:  This same talk will be given at the Cambridge Research
               Lab on Monday, 12/4, at 3:30PM.
	
	In this talk Dr. Lenat would like to present a surprisingly compact,
	powerful, elegant set of reasoning methods that form a set of first
	principles, which explain creativity, humor and common sense reasoning
	-- a sort of "Maxwell's Equations" of Thought.  He'd like very much
	to present them, but, sadly, he doesn't believe they exist.  So, 
	instead, he'll tell us what he's been working on down in Austin for
	the last five years.

	Intelligent behavior, especially in unexpected situations, requires
	being able to fall back on general knowledge, and being able to
	analogize to specific but far-flung knowledge.  But in order to
	learn something, one must already "almost know it".  I.e., learning
	occurs at the fringes of what we already know.  

	Unfortunately, the flip side of that comes into play every time we 
	build and run a program that doesn't know too much to begin with, 
	especially for tasks like semantic disambiguation of sentences, or 
	open-ended learning by analogy.  Expert systems finesse this by 
	restricting their tasks so much that they can perform relatively 
	narrow symbol manipulations which nevertheless are interpreted 
	meaningfully (and usefully) by human users.  But such systems are 
	hopelessly brittle:  they do not cope well with novelty, nor do they 
	communicate well with each other.

	OK, so the mattress in the road to AI is Lack of Knowledge.  But, how
	much does a program need to know, to begin with, in order to be non-
	brittle?  The annoying, inelegant, but apparently true answer is: a
	non-trivial fraction of consensus reality -- the tens of millions of
	things that we all know, and that we assume everyone else knows.  If
	a person were to liken the Stock Market to a roller-coaster, and you 
	didn't know what this means, they might liken it to a seesaw, or a 
	steel spring.  If you still didn't know what this means, they probably
	wouldn't want to deal with you anymore.





	It will take about two person-centuries to build up tha KB, assuming
	that we don't get stuck too badly on representation thorns along the
	way.  CYC -- Dr. Lenat's 1984-1994 project at MCC -- is an attempt to
	build that KB.  

	Some of the interesting issues are:  how they decide what knowledge 
	to encode, and how they encode it; how they represent substances, 
	parts, structures, time, space, belief, awareness and counterfactuals;
	how CYC can access, compute, inherit, deduce, or guess answers; how it
	computes and maintains plausibility and justification; and how they're
	coordinating a large team of knowledge enterers without having their
	semantics "diverge".

	They've gotten pretty far along already; in particular, they've found
	a tool-kit sort of answer for each of those problems:  a set of partial
	solutions which together cover most of the common situations.  

	This talk will present some of those solutions, and disucss the long-
	range (now just mid-range) impact of CYC:  (1) expert systems which
	understand the meaning of their terms, are less brittle, can inter-
	communicate, and reason via partial (analogical) matching; (2) natural
	language systems that can semantically disambiguate word senses, 
	resolve pronominal references, and induce the meaning of ellipses;
	and (3) machine learning systems that have a broad enough "fringe" 	
	to have some hope of solving problems by analogy.

	Today, no one would dream of buying a computer that lacked an operating
	system, or which couldn't support some spreadsheet and word processing
	applications.  In the late 90's, it is expected that no one would dream
	of buying a computer that lacked common sense.  Looking even farther
	ahead, it is foreseen that a Knowledge Utility will arise, not unlike
	the electric and telephone utilities in structure, scope, and impact
	on business and personal lives.

	Douglas Lenat's thesis, at Stanford in 1976, was a demonstration that
	certain kinds of creative discoveries in mathematics could be produced
	by a computer program.  This work earned him the biannual IJCAI Com-
	puters & Thought award in 1977.  

	He was named one of America's brightest scientists under the age of
	40 in the December, 1984, issue of "Science Digest".  (He's still under
	40, by the way, but just barely!)

	Dr. Lenat was a professor in the Computer Science Departments of CMU
	and Stanford prior to his move to MCC.  He remains a Consulting Pro-
	fessor and Industrial Lecturer at Stanford today.

	He is an editor for "IEEE Transactions on Knowledge and Data Engi-	
	neering", and the "Machine Learning Journal", and was co-founder of
	Teknowledge, Inc.  His main research interest is getting machines to
	discover new knowledge.  This will require, and has led him to work
	on, a large common sense knowledge base.






	Dr. Lenat has authored and co-authored more than 50 published papers
	and has written and edited several books, including:

	Knowledge Based Systems in Artificial Intelligence
	--------------------------------------------------
	
	Building Expert Systems
	-----------------------
	
	Building Large Knowledge-Based Systems:  Representation and Inference 
	---------------------------------------------------------------------
	in the CYC Project
	------------------
	
			 	 

67.48YESHERON::ROACHTANSTAAFL !Thu Nov 23 1989 12:29100
Printed by: Pat Roach                                    Document Number: 009228
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Nov-1989 01:33am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. MUSEN, 12/11, 1PM







	TITLE:		"Automated Generation of Custom-Tailored
			 Knowledge-Acquisition Tools"

	SPEAKER:	DR. MARK A. MUSEN
			Assistant Professor of Medicine 
			and Computer Science
			Stanford University

	DATE:		Monday, December 11, 1989

	TIME:		1:00-2:30PM

	PLACE:		Pacific Room DLB12

	HOST:		GEORG KLINKER
			Principal Software Engineer
			AI Research Group
			AI Technology Center


	Building a knowledge-based system is like developing a scientific
	theory.  Although a knowledge base does not constitute a theory 
	of some natural phenomenon, it does represent a theory of how a
	class of professionals approaches an application task.  As when 
	scientists develop a natural theory, builders of expert systems 
	first must formulate a model of the behavior that they wish to
	understand, and then must corroborate and extend that model with
	the aid of specific examples.  

	There are thus two interrelated phases of knowledge-base construc-
	tion:  (1) model building and (2) model extension.  Computer-based 
	tools can assist developers with both phases of the knowledge-acqui-
	sition process.  Workers in the area of knowledge acquisition have 
	developed computer-based tools that emphasize either the building 
	of new models or the extension of existing models.
                 
	The PROT�G� knowledge-acquisition system addresses these two 
	activities individually, and facilitates the construction of
	expert systems when the same general model can be applied to 
	a variety of application tasks.  

	Knowledge engineers work with application specialists to build 
	a model of a particular task area; PROT�G� then generates 
	automatically a graphical knowledge editor that is custom 
	tailored to the particular task area.  

	A nonprogrammer can use the PROT�G�-generated knowledge editor 
	without assistance to extend the task model entered into PROT�G�.
	This divide-and-conquer approach accelerates knowledge-base 	
	construction in application areas where multiple expert systems
	are required for related domain tasks.






	Dr. Mark Musen
	December 11, 1989
	Page Two



	Dr. Mark A. Musen received Sc.B. and M.D. degrees from Brown
	University in 1977 and 1980, respectively, and then pursued
	clinical training in Internal Medicine at Stanford University
	Hospital from 1980 to 1983.  He was certified by the American
	Board of Internal Medicine in 1983, and received the Ph.D. 
	degree from Stanford in 1988.  

	Dr. Musen was the recipient of the 1989 Young Investigator Award
	for Research in Medical Knowledge Systems from the American
	Association for Medical Systems and Informatics.

	His current research concentrates on the development of medical 
	expert systems, and on the design of automated tools that 
	facilitate the construction and maintenance of large knowledge 
	bases.  
67.49Managing Knowledge-Based SystemsHERON::ROACHTANSTAAFL !Thu Nov 23 1989 14:13113
Printed by: Pat Roach                                    Document Number: 009190
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Nov-1989 07:56am CET
                                        From:     ROACH
                                                  ROACH@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Managing Knowledge-Based Systems

	It is the hope of the AI Training and Education group to provide
	the many managers of the AITC with an opportunity to explore all 
	facets of managing, developing, and implementation of knowledge-
	base systems technology through their participation in the 
	Management Curriculum.  Our next offering is January 15 - 18, 1990.
	
	Please take a moment to read the following write-ups on the curriculum 
	to see if you and your group can benefit from the knowledge gained 
	from this course.

	AITE's Intro to AI in Business, or its equivalent, is a pre-requisite 
	to the management course.  Please note: the next Intro to AI in 
	Business is offered December 4 - 7th.

	To register, please call Suegene Levin at 291-8600 or send mail 
	to AIADM::Levin.





		     Management Curriculum Overview

The purpose of the Management Curriculum is to provide opportunities for
managers to explore the implications and issues in managing the development 
and implementation of knowledge-based systems technology.

The curriculum is composed of two (2) modules:

		INTRODUCTION TO AI IN BUSINESS
		MANAGING KNOWLEDGE-BASED SYSTEMS

INTRODUCTION TO AI IN BUSINESS is also the current first week of the Knowledge
Engineering curriculum and provides an overview of AI/knowledge-based systems 
technology.  It is open to both knowledge engineers and managers.

MANAGING KNOWLEDGE-BASED SYSTEMS focuses on the management implications of 
strategic, technical and organizational decisions in developing and implemen-
ting knowledge-based  systems.  It is a highly participative module, and 
requires that people bring an identified business problems(s) that could 
potentially use knowledge based systems as a solution.  Also, participants 
have opportunities to discuss plans and strategies for their back-home
applications with experienced practitioners.

A PRE-REQUISITE for the Managing Knowledge Based Systems is the Introduction 
to AI in Business course or its equivalent.

The AUDIENCE for this curriculum are those managers who manage knowledge 
engineers and/or the development, implementation, and/or use of knowledge-
based systems technology in their organizations.

FOR FURTHER INFORMATION:  Please refer to the course descriptions and schedule.

		



       	            MANAGING KNOWLEDGE-BASED SYSTEMS

CURRICULUM:  	Management

DESCRIPTION:   	The purpose of this course  is to enable managers to 
		identify the issues and plans for successfully introducing 
		and implementing knowledge based systems in their organiza-
		tions. It is a highly participative format with emphasis on 
		the individual's back-home application.  The content focuses 
		on the impact on the technology of strategic/business, techni-
		cal, human resource and organizational issues and decisions.  
		Opportunities for consultation on back-home issues are 
		provided throughout the program.  Participants work in small 
		groups with a total class size of not more than approx. 15 
		people.

GOALS:   	Upon completion of the course, the participants will be able 
		to:

		o Identify issues and plans for expert systems 
      		  applications in their own organizations.
		o Understand timing and justification strategies for 
		  a knowledge based systems project.
		o Identify staffing strategies using knowledge of the 
		  critical roles and skills required for success such 
		  as a knowledge engineer, program manager, and sponsor.
		o Understand issues on the host organization's readiness 
		  for the knowledge-based system, and how to manage 
		  expectations.

LENGTH 		4 days

FORMAT: 	Lecture/Discussion/Case Study/Consultation

PREREQUISITES:  Completion of INTRODUCTION OF AI IN BUSINESS or equivalent.  
		Participants must also have an identified problem(s) for 
		which an expert system is a potential solution.

	
		
67.50SYMBIOTICS, 12/14, 9:30AMHERON::ROACHTANSTAAFL !Wed Nov 29 1989 15:23116
Printed by: Pat Roach                                    Document Number: 009289
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     28-Nov-1989 03:53am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: SYMBIOTICS, 12/14, 9:30AM







	TITLE:		"Development Tools for Distributed Systems"

	SPEAKER:	DR. RICHARD ADLER
			Director of AI Research & Development
			Symbiotics, Inc.

	DATE:		Thursday, December 14, 1989

	TIME:		9:30-11:00AM

	NOTE:		A demonstration of the distributed communications 
			tool, MetaCourier, will follow the talk.

	PLACE:		Pacific Room, DLB12

	HOST:		DINO LACHIUSA
			Third Party Product Manager
			AI Marketing; AI Technology Center

	Distributed AI (DAI) deals with the cooperative solution of complex
	problems by a collection of (semi-)autonomous, intelligent problem-
	solving agents.  Problem data and control of problem-solving pro-
	cesses are distributed both logically and physically.  Agents 
	communicate through a message-passing model.  

	The primary focus of research to date in DAI has been on domains in 
	which an organization of distributed agents cooperates to solve a 
	single complex problem such as situation assessment.  The central 
	design issue is to develop fine-grained coordination mechanisms that 
	enable the agents to converge (efficiently) to a globally consistent 
	solution.  

	In contrast, "multiple problem" DAI systems involve multiple agents 
	and agent organizations that address multiple problems (e.g., config-
	uration, operation, fault detection and recovery), relating to a 
	common domain (e.g., support of a complex network system).  Critical 
	design issues here involve coarse-grained coupling of heterogeneous 
	agents, agent organizations, knowledge bases and data bases across 
	heterogeneous processing platforms.

	Work in progress on SOCIAL, a generalized framework for developing
	BOTH single and multiple problem DAI applications, will be described.
	
	SOCIAL provides developers with a set o generic agent and agent
	organization object classes, which can be instantiated and embedded
	with application elements.  These objects are built up by combining
	constructs from three object-oriented languages, which are also 
	directly accessible to developers through a second development
	interface.  





	Symbiotics, Inc. - 12/14/89 - PAGE TWO


	The MetaCourier language supplies the underlying functionality for 
	interprocess communication and control access across heterogeneous 
	computing environments (i.e., peer-to-peer movement of data between 
	application elements).  

	The MetaAgents language defines models for agent and agent organiza-
	tion coordination, control, message and resource management and 
	application-specific fault tolerance (e.g., blackboard, reception-
	ists).  

	The MetaViews language provides a common object-oriented data model 
	for peer-to-peer movement of data between heterogeneous data models 
	(e.g., different databases, database management systems).  

	Both of these languages sit on top of MetaCourier, which provides a 
	transparent distributed communications substrate for distributed 
	control and reasoning.

	This architecture conceals implementation details of communications,
	control and integration in distributed processing environments, 
	enabling developers to concentrate on the design and functionality 
	of the intelligent agents and agent organizations themselves.

	Dr. Richard Adler received his doctorate in Philosophy of Psysics
	from the Univ. of Minnesota, MS in Physics at the Univ.	of Illinois 
	at Urbana, and BS in Physics and Philosophy from the Univ. of Michigan.

	Prior to joining Symbiotics, Dr. Adler designed and implemented the 
	Operations Analyst (OPERA) system for NASA at the Kennedy Space Center.
 	OPERA consists of an integrated set of expert systems, implemented 
	using a hierarchical distributed blackboard architecture, that assist 
	in operational support of the primary control subsystem of the Space 
	Shuttle Launch Processing System.

	Dr. Adler's current research interests include DAI, symbolic simu-
	lation and temporal and causal reasoning.
		
67.51AITC TSS CALENDARHERON::ROACHTANSTAAFL !Wed Nov 29 1989 15:4537
Printed by: Pat Roach                                    Document Number: 009314
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     29-Nov-1989 05:00am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AITC TSS CALENDAR

				    DECEMBER
				    --------


	DATE/TIME/PLACE		SPEAKER/AFFILIATION	TITLE
	---------------         -------------------	-----

	12/5, Tuesday		Dr. Doug Lenat		"AI Versus
	9:30-11:00AM		MCC			Common Sense"
	DLB12, Pacific Room	


	12/11, Monday		Dr. Mark Musen		"Automated Generation
	1:00-2:30PM		Stanford University	of Custom-Tailored
	DLB12, Pacific Room				Knowledge-Acquisition
							Tools"


	12/14, Thursday		Dr. Richard Adler	"Development Tools
	9:30-11:00AM		Symbiotics, Inc.	for Distributed
	DLB12, Pacific Room				Systems"
67.52Dynamics of Difference OfferingHERON::ROACHTANSTAAFL !Mon Dec 04 1989 09:2784
Printed by: Pat Roach                                    Document Number: 009360
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     01-Dec-1989 08:08pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Dynamics of Difference Offering

 
                 UNDERSTANDING THE DYNAMICS OF DIFFERENCE


         COURSE NUMBER:   UDD0001
         DATES:           January 4-5,  or   March 29-30, 1990
         INSTRUCTOR:      Andrew Aull        Sandy Myers
         TIME:            8:30 - 5:00
         LOCATION:        Headmaster's House/W.Boylston, MA.
         COST:            $625

------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------

CLASS SIZE: 16

AUDIENCE:

U.D.D. is open to all employees in Distributed Systems.
Enrollment will be managed to maximize diversity in the course.
The course date will depend on the enrollment.

DRESS: Casual

CANCELLATION POLICY:
Must cancel 15 working days prior to course, or find a substitute;
otherwise your cost center will be charged.

COURSE GOALS:

Understanding the Dynamics of Difference provides a foundation for 
realizing the corporate Valuing Difference philosophy.  It will
help participants to:	

	o  recognize differences
        o  examine assumptions  
        o  explore stereotypes held about groups of people,
           organizations, functions, geographic areas, etc.
        o  seek out diversity and value differences as assets
        o  understand the significance of a value for difference 
           as a major variable in Digital's continued profitability
           and productivity

COURSE DESCRIPTION:

Understanding the Dynamics of Difference (UDD) is a two-day workshop
which provides an overview of the approach individuals use to develop
perceptions and assumptions.  It gives participants a process by
which to analyze the impact and/or results of these perceptions and
assumptions and provides the opportunity to slow down the Digital 
pace for reflective self-awareness.  Participants are guided to
examine their individual style and increase ease of interaction with
differences.

UDD is a course about valuing ourselves, as well as others who are
different from us. Participants are encouraged to consider how their
interactions with others may demonstrate the acceptance and valuing 
of others through individual and group exercises, videotapes and 
role plays.

Understanding the Dynamics of Difference helps participants to begin
or continue this process in a supportive environment.  For some
individuals, UDD provides a context in which to examine their current
level of commitment; for others it facilitates and expands their work
in valuing differences. UDD helps participants to see that differences
are assets, both personally in their own career and to the corporation.
67.53dR. LENAT, 12/5, 9:30AMHERON::ROACHTANSTAAFL !Wed Dec 06 1989 15:10152
Printed by: Pat Roach                                    Document Number: 009387
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     04-Dec-1989 11:48pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. LENAT, 12/5, 9:30AM





	TITLE:		"AI Versus Common Sense"

	SPEAKER:	DR. DOUGLAS B. LENAT
			Principal Scientist and Director
			MCC AI Laboratory

	DATE:		Tuesday, December 5, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room, DLB12

	HOSTS:		DR. JOHN MCDERMOTT, Manager, AI Research Group
			AI Technology Center, and
			BILL KUHLMAN, MCC Corporate Research Liaison


	Note:  This same talk will be given at the Cambridge Research
               Lab on Monday, 12/4, at 3:30PM.
	
	In this talk Dr. Lenat would like to present a surprisingly compact,
	powerful, elegant set of reasoning methods that form a set of first
	principles, which explain creativity, humor and common sense reasoning
	-- a sort of "Maxwell's Equations" of Thought.  He'd like very much
	to present them, but, sadly, he doesn't believe they exist.  So, 
	instead, he'll tell us what he's been working on down in Austin for
	the last five years.

	Intelligent behavior, especially in unexpected situations, requires
	being able to fall back on general knowledge, and being able to
	analogize to specific but far-flung knowledge.  But in order to
	learn something, one must already "almost know it".  I.e., learning
	occurs at the fringes of what we already know.  

	Unfortunately, the flip side of that comes into play every time we 
	build and run a program that doesn't know too much to begin with, 
	especially for tasks like semantic disambiguation of sentences, or 
	open-ended learning by analogy.  Expert systems finesse this by 
	restricting their tasks so much that they can perform relatively 
	narrow symbol manipulations which nevertheless are interpreted 
	meaningfully (and usefully) by human users.  But such systems are 
	hopelessly brittle:  they do not cope well with novelty, nor do they 
	communicate well with each other.

	OK, so the mattress in the road to AI is Lack of Knowledge.  But, how
	much does a program need to know, to begin with, in order to be non-
	brittle?  The annoying, inelegant, but apparently true answer is: a
	non-trivial fraction of consensus reality -- the tens of millions of
	things that we all know, and that we assume everyone else knows.  If
	a person were to liken the Stock Market to a roller-coaster, and you 
	didn't know what this means, they might liken it to a seesaw, or a 
	steel spring.  If you still didn't know what this means, they probably
	wouldn't want to deal with you anymore.





	It will take about two person-centuries to build up tha KB, assuming
	that we don't get stuck too badly on representation thorns along the
	way.  CYC -- Dr. Lenat's 1984-1994 project at MCC -- is an attempt to
	build that KB.  

	Some of the interesting issues are:  how they decide what knowledge 
	to encode, and how they encode it; how they represent substances, 
	parts, structures, time, space, belief, awareness and counterfactuals;
	how CYC can access, compute, inherit, deduce, or guess answers; how it
	computes and maintains plausibility and justification; and how they're
	coordinating a large team of knowledge enterers without having their
	semantics "diverge".

	They've gotten pretty far along already; in particular, they've found
	a tool-kit sort of answer for each of those problems:  a set of partial
	solutions which together cover most of the common situations.  

	This talk will present some of those solutions, and disucss the long-
	range (now just mid-range) impact of CYC:  (1) expert systems which
	understand the meaning of their terms, are less brittle, can inter-
	communicate, and reason via partial (analogical) matching; (2) natural
	language systems that can semantically disambiguate word senses, 
	resolve pronominal references, and induce the meaning of ellipses;
	and (3) machine learning systems that have a broad enough "fringe" 	
	to have some hope of solving problems by analogy.

	Today, no one would dream of buying a computer that lacked an operating
	system, or which couldn't support some spreadsheet and word processing
	applications.  In the late 90's, it is expected that no one would dream
	of buying a computer that lacked common sense.  Looking even farther
	ahead, it is foreseen that a Knowledge Utility will arise, not unlike
	the electric and telephone utilities in structure, scope, and impact
	on business and personal lives.

	Douglas Lenat's thesis, at Stanford in 1976, was a demonstration that
	certain kinds of creative discoveries in mathematics could be produced
	by a computer program.  This work earned him the biannual IJCAI Com-
	puters & Thought award in 1977.  

	He was named one of America's brightest scientists under the age of
	40 in the December, 1984, issue of "Science Digest".  (He's still under
	40, by the way, but just barely!)

	Dr. Lenat was a professor in the Computer Science Departments of CMU
	and Stanford prior to his move to MCC.  He remains a Consulting Pro-
	fessor and Industrial Lecturer at Stanford today.

	He is an editor for "IEEE Transactions on Knowledge and Data Engi-	
	neering", and the "Machine Learning Journal", and was co-founder of
	Teknowledge, Inc.  His main research interest is getting machines to
	discover new knowledge.  This will require, and has led him to work
	on, a large common sense knowledge base.






	Dr. Lenat has authored and co-authored more than 50 published papers
	and has written and edited several books, including:

	Knowledge Based Systems in Artificial Intelligence
	--------------------------------------------------
	
	Building Expert Systems
	-----------------------
	
	Building Large Knowledge-Based Systems:  Representation and Inference 
	---------------------------------------------------------------------
	in the CYC Project
	------------------
	
			 	 

67.54TSS at CRL Re. PROLOGHERON::ROACHTANSTAAFL !Wed Dec 06 1989 15:2690
Printed by: Pat Roach                                    Document Number: 009413
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     06-Dec-1989 00:16am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: TSS at CRL Re. PROLOG


 
 
		       Digital Equipment Corporation
			   Cambridge Research Lab
				  Seminar
 
		       Thursday, December 14th, 1989
				  10:00am
			   (Refreshments at 9:45)
 
  "Exploiting OR-Parallelism in Prolog using Multiple Sequential Machines"
 
			     Carole Susan Klein
			      Wolfson College
			  University of Cambridge
 
			        Abstract
 
 
	If the branches at each node of a tree are labelled, paths
	through the tree can be represented by a sequence of labels
	called an oracle.  If an oracle leading to a node is followed,
	all of the bindings and other state information associated with
	the node will be recreated.  Thus, oracles are both a
	specification for a path through the tree and a concise format
	for representing the environment at a particular node.
 
	This dissertation investigates the use of oracles for the
	parallel execution of Prolog programs.  The execution of a
	Prolog program can be represented pictorially by an AND/OR
	tree.  The branches of OR nodes within this tree have no
	binding dependencies so their evaluation can be performed on
	separate processors.  If one or more of these OR branches is
	explored in parallel, OR-parallelism is exploited in the Prolog
	program.
 
	A distributed system called the Delphi Machine has been
	designed and implemented to exploit the OR-parallelism inherent
	in Prolog programs.  In the implementation described in this
	dissertation, Delphi runs on a group of uniprocessors connected
	by Ethernet.  Various control strategies using oracles to
	control the parallel search are investigated.  The execution
	times for Prolog programs run on the Delphi machine are
	compared with those of a compiled and an interpreted sequential
	Prolog system.  The results show that a distributed system
	using oracles to control the parallel search can be an
	efficient way to exploit OR-parallelism in nondeterministic
	programs.
 
	Because of overheads imposed by the Delphi algorithm, a program
	executed on a single processor Delphi machine runs at
	approximately one half the speed as the same program executed
	on the unmodified Prolog system.  For a twenty processor
	configuration, the speed ups obtained vary from approximately
	two to nine times depending on the amount of OR-parallelism
	which can be exploited by Delphi.  Problems with large amounts
	of OR-parallelism show a nearly linear speed up.
 
------------------------------------------------------------------------------
 
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and
Broadway.  The address is:
 
                       Digital Equipment Corporation
                       Cambridge Research Lab
                       One Kendall Square, Bldg. 700
                       Cambridge, Mass 02139
 
For questions regarding the seminar, directions, or parking, or to add
your name to the mailing list, call DTN 259-6601 or send email to 
crl::tss.
 
67.55DR. MUSEN, 12/11, 1PMHERON::ROACHTANSTAAFL !Mon Dec 11 1989 11:09100
Printed by: Pat Roach                                    Document Number: 009490
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     09-Dec-1989 00:11am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@STATOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. MUSEN, 12/11, 1PM







	TITLE:		"Automated Generation of Custom-Tailored
			 Knowledge-Acquisition Tools"

	SPEAKER:	DR. MARK A. MUSEN
			Assistant Professor of Medicine 
			and Computer Science
			Stanford University

	DATE:		Monday, December 11, 1989

	TIME:		1:00-2:30PM

	PLACE:		Pacific Room DLB12

	HOST:		GEORG KLINKER
			Principal Software Engineer
			AI Research Group
			AI Technology Center


	Building a knowledge-based system is like developing a scientific
	theory.  Although a knowledge base does not constitute a theory 
	of some natural phenomenon, it does represent a theory of how a
	class of professionals approaches an application task.  As when 
	scientists develop a natural theory, builders of expert systems 
	first must formulate a model of the behavior that they wish to
	understand, and then must corroborate and extend that model with
	the aid of specific examples.  

	There are thus two interrelated phases of knowledge-base construc-
	tion:  (1) model building and (2) model extension.  Computer-based 
	tools can assist developers with both phases of the knowledge-acqui-
	sition process.  Workers in the area of knowledge acquisition have 
	developed computer-based tools that emphasize either the building 
	of new models or the extension of existing models.
                 
	The PROT�G� knowledge-acquisition system addresses these two 
	activities individually, and facilitates the construction of
	expert systems when the same general model can be applied to 
	a variety of application tasks.  

	Knowledge engineers work with application specialists to build 
	a model of a particular task area; PROT�G� then generates 
	automatically a graphical knowledge editor that is custom 
	tailored to the particular task area.  

	A nonprogrammer can use the PROT�G�-generated knowledge editor 
	without assistance to extend the task model entered into PROT�G�.
	This divide-and-conquer approach accelerates knowledge-base 	
	construction in application areas where multiple expert systems
	are required for related domain tasks.






	Dr. Mark Musen
	December 11, 1989
	Page Two



	Dr. Mark A. Musen received Sc.B. and M.D. degrees from Brown
	University in 1977 and 1980, respectively, and then pursued
	clinical training in Internal Medicine at Stanford University
	Hospital from 1980 to 1983.  He was certified by the American
	Board of Internal Medicine in 1983, and received the Ph.D. 
	degree from Stanford in 1988.  

	Dr. Musen was the recipient of the 1989 Young Investigator Award
	for Research in Medical Knowledge Systems from the American
	Association for Medical Systems and Informatics.

	His current research concentrates on the development of medical 
	expert systems, and on the design of automated tools that 
	facilitate the construction and maintenance of large knowledge 
	bases.  
67.56SYMBIOTICS, 12/14, 9:30AMHERON::ROACHTANSTAAFL !Wed Dec 13 1989 10:29116
Printed by: Pat Roach                                    Document Number: 009555
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Dec-1989 05:16am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@MRGATE@DELOS@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: SYMBIOTICS, 12/14, 9:30AM







	TITLE:		"Development Tools for Distributed Systems"

	SPEAKER:	DR. RICHARD ADLER
			Director of AI Research & Development
			Symbiotics, Inc.

	DATE:		Thursday, December 14, 1989

	TIME:		9:30-11:00AM

	PLACE:		Pacific Room, DLB12

	HOST:		DINO LACHIUSA
			Third Party Product Manager
			AI Marketing; AI Technology Center

	**NOTE:	THERE WILL *NOT* BE A DEMONSTRATION OF THIS
                TOOL AT THIS TIME, AS ORIGINALLY ADVERTISED.
	
	Distributed AI (DAI) deals with the cooperative solution of complex
	problems by a collection of (semi-)autonomous, intelligent problem-
	solving agents.  Problem data and control of problem-solving pro-
	cesses are distributed both logically and physically.  Agents 
	communicate through a message-passing model.  

	The primary focus of research to date in DAI has been on domains in 
	which an organization of distributed agents cooperates to solve a 
	single complex problem such as situation assessment.  The central 
	design issue is to develop fine-grained coordination mechanisms that 
	enable the agents to converge (efficiently) to a globally consistent 
	solution.  

	In contrast, "multiple problem" DAI systems involve multiple agents 
	and agent organizations that address multiple problems (e.g., config-
	uration, operation, fault detection and recovery), relating to a 
	common domain (e.g., support of a complex network system).  Critical 
	design issues here involve coarse-grained coupling of heterogeneous 
	agents, agent organizations, knowledge bases and data bases across 
	heterogeneous processing platforms.

	Work in progress on SOCIAL, a generalized framework for developing
	BOTH single and multiple problem DAI applications, will be described.
	
	SOCIAL provides developers with a set o generic agent and agent
	organization object classes, which can be instantiated and embedded
	with application elements.  These objects are built up by combining
	constructs from three object-oriented languages, which are also 
	directly accessible to developers through a second development
	interface.  





	Symbiotics, Inc. - 12/14/89 - PAGE TWO


	The MetaCourier language supplies the underlying functionality for 
	interprocess communication and control access across heterogeneous 
	computing environments (i.e., peer-to-peer movement of data between 
	application elements).  

	The MetaAgents language defines models for agent and agent organiza-
	tion coordination, control, message and resource management and 
	application-specific fault tolerance (e.g., blackboard, reception-
	ists).  

	The MetaViews language provides a common object-oriented data model 
	for peer-to-peer movement of data between heterogeneous data models 
	(e.g., different databases, database management systems).  

	Both of these languages sit on top of MetaCourier, which provides a 
	transparent distributed communications substrate for distributed 
	control and reasoning.

	This architecture conceals implementation details of communications,
	control and integration in distributed processing environments, 
	enabling developers to concentrate on the design and functionality 
	of the intelligent agents and agent organizations themselves.

	Dr. Richard Adler received his doctorate in Philosophy of Psysics
	from the Univ. of Minnesota, MS in Physics at the Univ.	of Illinois 
	at Urbana, and BS in Physics and Philosophy from the Univ. of Michigan.

	Prior to joining Symbiotics, Dr. Adler designed and implemented the 
	Operations Analyst (OPERA) system for NASA at the Kennedy Space Center.
 	OPERA consists of an integrated set of expert systems, implemented 
	using a hierarchical distributed blackboard architecture, that assist 
	in operational support of the primary control subsystem of the Space 
	Shuttle Launch Processing System.

	Dr. Adler's current research interests include DAI, symbolic simu-
	lation and temporal and causal reasoning.
		
67.57Object-Oriented Tech. OfferingHERON::ROACHTANSTAAFL !Mon Jan 08 1990 08:1866
Printed by: Pat Roach                                    Document Number: 009808
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     06-Jan-1990 07:26am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Object-Oriented Tech. Offering

********************************************************************************
           OBJECT-ORIENTED TECHNOLOGY AND ITS ROLE IN DATABASE SYSTEMS
********************************************************************************

  OFFERINGS:   January 22 - 26th, 1990
	       June 12 - 16th, 1990

   LOCATION:   DLB12-1, Marlborough, MA

DESCRIPTION:   This course will focus on the area of object-oriented databases.
               Object-oriented databases have been proposed as a promising tech-
	       nology for storing complex data in applications like programming
	       environments, CAD, or knowledge databases.  In order to fully
	       appreciate this new technology, this course will begin by exa-
	       mining the principles of object-oriented programming and conven-
	       tional databases.  In the area of object-oriented programming,
	       we will look at the ways in which basic issues of software eng-
	       ineering are addressed by means of strong typing, abstraction,
	       and inheritance.  In the area of conventional databases, we will
	       look at relational languages and transaction processing schemes.
	       In both of these areas, the principles will be illustrated with
               examples from real systems.

	       The emphasis of the course will be on object-oriented databases.
	       We will look at some specific proposals in this area.  Some of
	       the issues that we will address include data models, system 
	       architectures, version control, schema evolution, complex 
	       objects, and cooperative design transactions.  In this section,
	       the class will participate in an object-oriented database design.
	       Guest speakers will describe two on-going projects within DEC
	       that are involved in object-oriented databases.  One is the 
	       design and implementation of an object-oriented database system,
	       and the other is a CAD application that is using an object-
	       oriented database system as to store its data.

PREREQUISITES: Participants are assumed to have experience with large-scale
	       system building efforts.  They should be familiar with the
	       fundamentals of computer science like languages design,
	       operating systems, and data structures.
		 
  PRESENTER:   Dr. Stanley Zdonik, Assistant Professor, Brown University

     FORMAT:   Lectures, five days

       COST:   $2,500. 

   REGISTER:   Send mail to AIADM::KING.  Please include badge number, DTN,
	       cost center, group name, organization name and manager's name.
       
CANCELLATION:  10 Days before start of class
67.58YESHERON::ROACHTANSTAAFL !Fri Jan 12 1990 18:31102
Printed by: Pat Roach                                    Document Number: 009894
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Jan-1990 05:51pm CET
                                        From:     CUMMINGS
                                                  CUMMINGS@GUESS@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject:  



	

+---------------------------+
|   |   |   |   |   |   |   |
| d | i | g | i | t | a | l |		I N T E R O F F I C E  M E M O
|   |   |   |   |   |   |   |
+---------------------------+


TO:  @AITC.DIS		                DATE: 15 January 1990
					FROM: Chuck Marshall	 
					DEPT: ASTG 
					 EXT: 291-8217
					L/MS: DLB5-3/A10 
					NODE: GUESS::Marshall 


SUBJECT: Innovative Design of Mechanical Structures from First Principles


On Monday January 15th Jon Cagan, University of California at Berkeley, will 
be here to give a presentation on Innovative Design of Mechanical Structures 
from First Principles.  This presentation is scheduled from 2:00-3:00 in 
Fur Seals conference room (DLB5-3/D10).   If you are interested in 
attending please notify Gail Cummings, GUESS::CUMMINGS or DTN 291-8943, 
to confirm your attendance.   Attached is a brief abstract.



  


             Innovative Design of Mechanical Structures from First
                               Principles


                             Jonathan Cagan
                   Intelligent Systems Research Group
                    Mechanical Engineering Department
                  University of California at Berkeley
    
    
    
    Recent interest in design research has been motivated by the develop-
    ment of expert system tools which permit computational reasoning.  
    However, expert  systems tend to be shallow and are best applied to 
    routine levels of design; once the scope of a problem falls outside 
    the domain of the expert system, the program is likely to give inap-
    propriate responses.
    
    
    This presentation discusses a non-routine design methodology called 1st 
    PRINCE which reasons from a deep level of knowledge for the mechanical 
    structures domain.  The method is based on the assumption that the creation
    of innovative designs of physical significance, concerning geometric and 
    material properties, requires reasoning from first principles.  The inno-
    vative designs discovered by 1st PRINCE differ from routine designs in that
    new primitives are created.  Monotonicty  analysis and computer algebra are
    utilized to direct design variables in a globally optimal direction rela-
    tive to the goals specified.  Expansion of the design configuration space 
    and the creation of new primitives, in order to meet the constraints or 
    improve the design, are achieved by manipulating mathematical quantities 
    such as division of the integral.  Inductive techniques observe trends in 
    the solution space to determine the optimally directed limit of the design 
    modifications.  The methodology emphasizes the importance of optimization 
    in the design problems.  A graph theoretic representation is developed to 
    provide an environment with which to perform non-routine design of physical
    structures.  Although the area of emphasis in this discussion is mechanical
    structures, the methodology itself is domain independent. 
    
    
    By minimizing weight, 1stPRINCE is a applied to a beam under torsion load 
    where hollow tubes and composite rods are discovered and to a beam under 
    flexural load where a tapered beam is innovated.  During application to a 
    block while minimizing its resistance to spinning, 1stPRINCE invents a 
    circular wheel. 








67.59AITC Quarterly meeting schedule revisionHERON::ROACHTANSTAAFL !Tue Jan 16 1990 20:0939
Printed by: Pat Roach                                    Document Number: 009966
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     16-Jan-1990 07:00pm CET
                                        From:     HUNT
                                                  HUNT@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AITC Quarterly meeting schedule revision

From:	AIADM::MARSHALL     "Barry Marshall - 291-8447" 16-JAN-1990 11:32:08.06
To:	HUNT,MARSHALL
CC:	
Subj:	revised agenda - please send out again, thanks



To: AITC employees

The AITC Quarterly meeting is scheduled for Thursday, January 18, 1990,
from 2-4 pm in the cafeteria at DLB 5.


	Rewards & Recognition


	Applied Intelligent Systems Group Overview - Mitch Tseng


	Productivity Shells Update - Jack Rahaim


	Organizational Update - Dennis O'Connor
67.60For your consideration - an advance course offeringHERON::ROACHTANSTAAFL !Thu Jan 18 1990 19:2787
Printed by: Pat Roach                                    Document Number: 009993
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     17-Jan-1990 09:35pm CET
                                        From:     KING
                                                  KING@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: For your consideration - an advance course offering

********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		February 8 - 9, 1990
		April 12 - 13, 1990

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Theresa Chow, Instructor

AUDIENCE and
PREREQUISITES:	The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1000 
    
REGISTER:	Send mail to AIADM::KING.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.61AITC TRAINING CALENDARHERON::ROACHTANSTAAFL !Tue Jan 23 1990 12:4756
Printed by: Pat Roach                                    Document Number: 010024
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Jan-1990 06:13pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AITC TRAINING CALENDAR

------------------------------------------------------------------------
To register:  Set Host DELNI; Username: Courses; Password: not required
	      Or contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------


        			AITC TRAINING CALENDAR


	         		   FY90 -- Q3/Q4


   DATE		     COURSE		  INSTRUCTOR	 BLDG--LOCATION--C.R.

FEBRUARY 26-27  ESSENTIALS OF             CHRIS MIKES      DLB5-3 / LEOPARD 
                NEGOTIATION

MARCH 1-2       EXECUTIVE		  EXECUTIVE	   DLB5-3 / LEOPARD
		PRESENTATIONS		  TECHNIQUE

MARCH 26-28     MANAGING PERFORMANCE      PETER GARDNER    DLB12-1 / ATLANTIC 

APRIL 9,11,18   LISTENING SKILLS	  TBD		   DLB5-3  / LEOPARD

APRIL 19-20     WRITING FOR RESULTS	  LINDA MARKS      DLB5-03 / LEOPARD

APRIL 25-27	COUNSELING SKILLS	  RONNIE MILLER    DLB5-03 / LEOPARD
                FOR MANAGERS

MAY 1-4         POSITIVE POWER &	  STEVE ISENBERG   DLB12-1 / PACIFIC
	        INFLUENCE

MAY 21		TIME MANAGEMENT		  FRANKLIN	   DLB12-1 / PACIFIC
	                                  INSTITUTE
     
JUNE 4	        COPING WITH STRESS        DIRK VENEMA	   DLB12-1 / PACIFIC


Please feel free to contact Faye Napert, DELNI::NAPERT or John LeBlanc,
AIADM::LEBLANC, if you have any questions or need additional information.
67.62TESTING K-B SYSTEMSHERON::ROACHTANSTAAFL !Tue Jan 23 1990 12:4890
Printed by: Pat Roach                                    Document Number: 010026
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Jan-1990 09:29pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: TESTING K-B SYSTEMS


	        Please send all inquiries to AIADM::King

********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		February 8 - 9, 1990
		April 12 - 13, 1990

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Theresa Chow, Instructor

AUDIENCE and
PREREQUISITES:	The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1000 
    
REGISTER:	Send mail to AIADM::KING.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.63EXEC. PRES. PROG. OFFERINGHERON::ROACHTANSTAAFL !Tue Jan 23 1990 12:5858
Printed by: Pat Roach                                    Document Number: 010042
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Jan-1990 09:07pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EXEC. PRES. PROG. OFFERING

                       EXECUTIVE PRESENTATIONS

=========================
DS/NaC EDUCATION SERVICES
=========================

         COURSE NUMBER:       EXE0001
         DATE:                MARCH 1-2, 1990
         TIME:                8:30 - 5:00
         LOCATION:            LEOPARD C/R DLB
         INSTRUCTOR:          TBA
         COST:                $600

------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------


Intended Audience:

Middle managers 

Course Description:

The Executive Presentations Program is designed for middle managers to help
them enhance their oral communication skills within small or large business
groups.  The intensive course focuses on the organization and delivery of
ideas in a convincing and persuasive manner.

The course uses a three-member team to provide personal and videotaped
feedback to each participant.  There is prework, and each participant is 
videotaped during the course, allowing time for individualized coaching, 
reinforcement of a major theme, and development of personal style.  

Course Objectives:

Participants will:

o  Organize ideas and communicate them clearly and persuasively under pressure
o  Convey a clear and succinct message
o  Create and use visual aids in making effective presentations
67.64ADVANCED OPS5 OFFERINGHERON::ROACHTANSTAAFL !Mon Jan 29 1990 09:5771
Printed by: Pat Roach                                    Document Number: 010104
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     26-Jan-1990 07:40pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ADVANCED OPS5 OFFERING

  TO REGISTER:  Send name, badge number, group name and cost center to 
	        AIADM::King

********************************************************************************
		   Advanced Course Curriculum Announcement
			       ADVANCED VAX OPS5
********************************************************************************

	 DATE:  February 5 - 9th, 1990

     LOCATION:  Lab 2, DLB12 - 295 Donald Lynch Boulevard, Marlborough, MA

  DESCRIPTION:  The Advanced VAX OPS5 Course takes a close look at topics of 
		concern to experienced OPS5 programmers.  The topics are 
		discussed both theoretically and practically, using significant 
		lab time and OPS5 programs, for examples.

		The lab exercises used to practice the concepts taught require
		modification of existing code or rapid generation of new code.
		The modules concerning external routines require that the 
		student understand simple programs in other VAX languages.

        GOALS:  The Advanced OPS5 course enables the participants to:

		  o Sharpen basic OPS5 programming skills, including
		    data representation and effective use of control.
	
		  o Practice OPS5 programming techniques, and understand 
                    the advantages and disadvantages of using those techniques.

		  o Encourage good OPS5 programming principles, and understand
		    why they are good.
		    
		  o Understand how the RETE match algorithm is implemented in
                    VAX OPS5, and how to write rules that maximize its 
		    efficient use.

	         o  Improve your understanding of integration and delivery
	            issues such as application architecture, testing and
	            evolution.

       LENGTH:  One week

       FORMAT:  Lecture/lab

PREREQUISITES:  Participants in this course should have at least 6 months
		experience programming in OPS5, including experience with a
		large project.  For modules on external routines, participants
		must be able to understand simple programs in a procedural VAX
		language that can communicate with OPS5.
       
   PRESENTERS:  Lesley Chesler, Tom Cooper

  TO REGISTER:  Send name, badge number, group name and cost center to 
	        AIADM::King
67.65BLACKBOARD ARCHITECTUREHERON::ROACHTANSTAAFL !Tue Jan 30 1990 10:1571
Printed by: Pat Roach                                    Document Number: 010128
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     30-Jan-1990 01:58am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: BLACKBOARD ARCHITECTURE

NOTE:	   Please send all replies to AIADM::King

********************************************************************************
	   AI Training and Education's Advanced Course Offering
                	BLACKBOARD-BASED AI SYSTEMS
*******************************************************************************

DATE:		March 14-16th, 1990

LOCATION:	Lab 2, Meditteranean Classroom
		DLB12-1

DESCRIPTION: 	The blackboard paradigm is a powerful and flexible
		problem-solving architecture.  This course provides 
		a detailed examination of the blackboard paradigm and 
		laboratory experience building a blackboard-based AI 
		application.  Emphasis is placed on the advantages/dis-
		advantages of the blackboard approach and on character-
		istics of applications that make them well-suited to
		blackboard technology.  Finally, the blackboard paradigm 
		provides a number of problem-solving control opportunities, 
		and we consider recent developments in blackboard control 
		and meta-level control mechanisms.

GOALS: 		Upon completion of the course, participants will have:

			o A working knowledge of how blackboard-based 
		          AI systems function and an understanding of 
			  what types of applications are appropriate
			  for blackboard architectures
			o A historical perspective on the development 
			  of blackboard technology over the last decade 
			  and a view of where research on blackboard 
			  approaches is headed in the coming years
			o An understanding of the efficiency and control 
			  issues that must be addressed in developing a 
			  high-performance blackboard-based application
			o Had the opportunity to build a small blackboard-
			  based scheduling application using the Generic 
			  Blackboard Development System (GBB)

LENGTH: 	3 day

FORMAT: 	Lecture/Lab

PREREQUISITES:  General AI knowledge, programming ability in LISP or OPS5,
		and knowledge of an editor

PRESENTER: 	Dr. Daniel Corkill
		Dr. Victor Lesser, UMASS Amherst

COST:		$2,000

REGISTRATION:	Send name, badge number, cost center, group/organizational
	        name and manager's name to AIADM::King 
67.66PROF. FEHLING, 2/6, 9:30AMHERON::ROACHTANSTAAFL !Tue Jan 30 1990 10:15116
Printed by: Pat Roach                                    Document Number: 010129
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     30-Jan-1990 04:43am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PROF. FEHLING, 2/6, 9:30AM






	TITLE:		"A Systems-and-Problem-Oriented Approach
			to Intelligent Process-Management"
 
	SPEAKER:        PROFESSOR MICHAEL FEHLING
		   	Director, Intelligent Systems Laboratory
			Stanford University
		        
	DATE:		Tuesday, February 6, 1990

	TIME:		9:30-11:00AM

	PLACE:		Leopard Conference Room
			DLB5-3/D9


	HOSTS:		MITCH TSENG, Group Manager and
			TOM CERVA, Planning Technical & Quality Manager
			Applied Intelligent Systems Group (AISG)
			AI Technology Center
	 


	The Intelligent Systems Laboratory (ISL) at Stanford is engaged in 
	research to create new technology for application to "process-manage-
	ment" tasks.  Examples of process management applications include 
	control of complex, and typically, distributed manufacturing processes,
	system-wide control of autonomous robotic vehicles, and automated 
	management of advanced aerospace systems.  Much of this research is 
	interdisciplinary, combining concepts and methods from AI and various 
	systems-disciplines such as control theory, optmization, and the 
	decision sciences.  Our research is also problem-oriented. We emphasize
	early application of new methods to practical, real-world problems.
 
	This talk will review some important aspects of our work on Intelligent
	Process Management.  We shall describe Schemer, an AI architecture that
	provides a generic framework for real-time, process-management applica-
	tions. To the best of our knowledge, Schemer is the only AI architec-
	ture specifically developed to support the implementation of distri-
	buted, real-time problem-solving systems.  For this reason, Schemer 
	provides a set of unique features that are well-suited for building 
	process-management applications. We shall describe these features and 
	the role they play in real-time problem-solving systems.  As time 
	permits we shall discuss some specific problem-solving techniques that
	we are developing within Schemer for use in process-management applica-
	tions.  We shall conclude our presentation by sketching some examples 
	of how results of our research have been rapidly and successfully 
	deployed and used to significantly enhance current manufacturing
	technology.







	"A Systems-and-Problem-Oriented Approach
	to Intelligent Process-Management"
	Professor Michael Fehling
	Page Two
 
 
	For over 15 years, Prof. Michael Fehling has been a major contributor 
	to basic and applied research in the fields of AI and the cognitive- 
	and decision-sciences.  His interdisciplinary research interests and 
	methodology have evolved as a result of his broad experience in both
	academic and industrial settings.  Before joining Stanford's faculty, 
	Prof. Fehling has held the positions of Principle Scientist of the 
	Rockwell International Science Center, Program Director and Senior 
	Scientist at Teknowledge, Inc. and Program Manager at Advanced Decision
	Systems.  He has also held a faculty position at Rutgers University.

	As Director of Stanford's Intelligent Systems Laboratory (ISL), Prof. 
	Fehling is continuing to lead interdisciplinary research on "resource-
	constrained problem-solving" by individual problem-solving agents as 
	well as by multiple, interacting agents. Prof. Fehling is investigating
	how intelligent problem-solving agents dynamically adapt their reason-
	ing and decision-making activities to satisfy contextually determined 
	constraints.  More specifically, this research is producing (1) new 
	methods for meeting real-time performance constraints in problem-
	solving applications and (2) practical methods for managing uncertainty.
  	These methods are being used to enhance both centralized and highly 
	distributed forms of problem-solving.  This interdisciplinary research
 	combines computational methods from AI and computer science with 
	quantitative, analytic approaches from the systems sciences.
 
	Prof. Fehling's work contributes significantly to the growth of new, 
	practical technology.  Prof. Fehling and his colleagues and students 
	developed and deployed many prototype and fully operational systems 
	for important applications such as industrial process control, 
	logistics and factory scheduling, and control of advanced aerospace 
	systems.  Prof. Fehling holds a patent for the development of CIM-
	FLEX-Teknowledge's ABE(TM) system, a new-generation software develop-
	ment environment for constructing large-scale, distributed systems 
	that combine AI and non-AI computational methods.  
 
67.67dAX DECISION EXPERTHERON::ROACHTANSTAAFL !Tue Feb 06 1990 14:3376
Printed by: Pat Roach                                    Document Number: 010245
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     02-Feb-1990 11:50pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: VAX DECISION EXPERT

     		-------------------------------------------------
    		=                                               =
    		=   Educational Services Intelligent Systems    = 
    		=                                               = 
    		=               DIGITAL EQUIPMENT               = 
    		=                                               = 
    		=                offers the course	        =
    		=                                               =  
    		=       PROGRAMMING IN VAX DECISION EXPERT      = 
    		=                                               =
    		-------------------------------------------------

        
    VAX Decision Expert is an Expert System Shell from Digital
    Equipment.  Educational Services offers a 5 day course on VAX
    Decision Expert. 
   
    Course Description: 
    ------------------ 
    The course is designed to train non-AI programmers and engineers in the
    use of Digital Equipment's new Expert System Development Tool VAX
    Decision Expert.
    
    The goal of the course is to cover the fundamentals of VAX Decision
    Expert as well as giving a basic understanding of Expert Systems,
    inference and rules in general. We will teach how to program and use
    the system, all about the knowledge representations, different types of
    inference methods and how they are used in VAX Decision Expert. The
    importance of utility language, how to use it and call-in, call-out
    features to access external programs and devices are also covered in
    the course as well as managing knowledge bases, modules and how to link
    modules together. Covered in the course is also how to build a
    deliverable application.
   
    Working on VAX stations, students will have the opportunity to develop
    practical skills using both VAX Decision Expert's development and
    delivery environments. Over the week the students will do exercises
    tightly coupled with the lectures, as well as developing a larger
    project.
   
     
    If you are interested in taking this course, please contact the
    Educational Services Course Registration Office in Bedford.
    
    Course Data: 
    ----------- 
    Phone number to registration office in Bedford: (617) 276-4380 
    Course name:    	Programming in VAX Decision Expert 
    Course number:  	EY-A957E-L0 
    Course dates:	February 19, 1990 
    Course length:  	5 days 
    Location:       	AI Technology Center, DLB12,  295 Donald
	    		Lynch Boulevard, Marlboro, MA    
    Price:		$1095

     	
    Information:
    -----------
    For more information about the VAX Decision Expert Course and future 
    course dates, please call Malin Goodwin at DTN:291-8894 or (508) 490-8894.
67.68RICH+BARNETT,2/9,9:30HERON::ROACHTANSTAAFL !Tue Feb 06 1990 14:34121
Printed by: Pat Roach                                    Document Number: 010246
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     03-Feb-1990 00:36am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: RICH+BARNETT,2/9,9:30


   




	TITLE:     	"Knowledge-Based Natural Language"

	SPEAKERS:  	JIM BARNETT, KBNL Group Leader, and

		   	DR. ELAINE RICH, Co-director
			Artificial Intelligence Laboratory
	   		Microelectronics and Computer Technology 
			Corporation (MCC)

	PLACE:     	Cafeteria, DLB5

	DATE:      	Friday, February 9, 1990

	TIME:      	9:30-11:00 A.M.

	HOST:      	BILL KUHLMAN
	   		MCC Liaison, Corporate Research




	Natural language and knowledge about the world go together.  This 
	is hardly a new observation.  But it is only beginning to be a  
	correct one from a computational point of view, because, although 
	it is true for the world as a whole and a language as a whole, it 
	is usually not true for tiny slices out of either of them.  Yet 
	the best computational models have been of only relatively tiny 
	slices, particularly of the world, and to a lesser extent, of  
	language.

	In this talk, we will describe our work on a Knowledge Based Natural 
	Language System, which relies on the CYC common sense knowledge base 
	that Doug Lenat is building at MCC.  

	A key idea behind this work is that the natural relationships that 
	exist between the way the world is structured and the way we use 
	language to talk about the world introduce redundancy into the 
	process of creating a knowledge base about the world and a parallel 
	knowledge base about a language.


	This redundancy can be used:  to simplify the task of creating a 
	KB about the world; to simplify the task of defining a lexicon for 
	NL processing; to detect inconsistencies in both the KB and the 
	lexicon; and to produce a resulting NL system that is far less 
	brittle than existing ones.






	"Knowledge-Based Natural Language"
	Jim Barnett and Dr. Elaine Rich
	Page Two



	The KBNL system is composed of five subsystems:

   		1. Lucy, an English understanding system

 	  	2. Koko, an English generation system

   		3. Tell, a knowledge-based lexical acquisition 
		   system that supports both Lucy and Koko

   		4. Show, a tool for retrieving objects from a 
	           knowledge base using English descriptions

   		5. Scan, a full text retrieval system based 
                   on Lucy and Koko.

	The various stages of development of these systems will also be 
	described.

	Jim Barnett is the group leader of the KBNL project at MCC, and has  
	been the major architect of the semantic interpretation component 
	of that system since joining the group several years ago.  

	He has a B.A. in History from Connecticut College, an M.S. in 
	Linguistics from Cornell University, and is currently pursuing a  
	Ph.D. in Linguistics at the Univertity of Texas, where his disser-
	tation work is focusing on disambiguation strategies.

	Elaine Rich is co-director of the Artificial Intelligence Lab in 
	MCC's Advanced Computing Technology (ACT) Program, where she is 
	currently responsible for two projects:  KBNL, a large-scale,  
	knowledge-based natural language processing effort; and RA, whose  
	goal is to produce a platform that supports the construction of  
	hybrid and distributed knowledge-based systems.  

	Prior to joining the AI lab, she was associate director of the  
	Human Interface Laboratory at MCC, where she developed an  
	architecture for intelligent, multimodal interfaces.  She was 
	an assistant professor of computer sciences at the  University  
	of Texas at Austin prior to joining MCC.  She received an A.B.
	in linguistics and applied mathematics from Brown University  
	and a Ph.D. in computer science from Carnegie-Mellon University.
67.69d/6,PROF. FEHLING,9:30AMHERON::ROACHTANSTAAFL !Tue Feb 06 1990 14:57116
Printed by: Pat Roach                                    Document Number: 010264
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     06-Feb-1990 00:27am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: 2/6,PROF. FEHLING,9:30AM






	TITLE:		"A Systems-and-Problem-Oriented Approach
			to Intelligent Process-Management"
 
	SPEAKER:        PROFESSOR MICHAEL FEHLING
		   	Director, Intelligent Systems Laboratory
			Stanford University
		        
	DATE:		Tuesday, February 6, 1990

	TIME:		9:30-11:00AM

	PLACE:		Leopard Conference Room
			DLB5-3/D9


	HOSTS:		MITCH TSENG, Group Manager and
			TOM CERVA, Planning Technical & Quality Manager
			Applied Intelligent Systems Group (AISG)
			AI Technology Center
	 


	The Intelligent Systems Laboratory (ISL) at Stanford is engaged in 
	research to create new technology for application to "process-manage-
	ment" tasks.  Examples of process management applications include 
	control of complex, and typically, distributed manufacturing processes,
	system-wide control of autonomous robotic vehicles, and automated 
	management of advanced aerospace systems.  Much of this research is 
	interdisciplinary, combining concepts and methods from AI and various 
	systems-disciplines such as control theory, optmization, and the 
	decision sciences.  Our research is also problem-oriented. We emphasize
	early application of new methods to practical, real-world problems.
 
	This talk will review some important aspects of our work on Intelligent
	Process Management.  We shall describe Schemer, an AI architecture that
	provides a generic framework for real-time, process-management applica-
	tions. To the best of our knowledge, Schemer is the only AI architec-
	ture specifically developed to support the implementation of distri-
	buted, real-time problem-solving systems.  For this reason, Schemer 
	provides a set of unique features that are well-suited for building 
	process-management applications. We shall describe these features and 
	the role they play in real-time problem-solving systems.  As time 
	permits we shall discuss some specific problem-solving techniques that
	we are developing within Schemer for use in process-management applica-
	tions.  We shall conclude our presentation by sketching some examples 
	of how results of our research have been rapidly and successfully 
	deployed and used to significantly enhance current manufacturing
	technology.







	"A Systems-and-Problem-Oriented Approach
	to Intelligent Process-Management"
	Professor Michael Fehling
	Page Two
 
 
	For over 15 years, Prof. Michael Fehling has been a major contributor 
	to basic and applied research in the fields of AI and the cognitive- 
	and decision-sciences.  His interdisciplinary research interests and 
	methodology have evolved as a result of his broad experience in both
	academic and industrial settings.  Before joining Stanford's faculty, 
	Prof. Fehling has held the positions of Principle Scientist of the 
	Rockwell International Science Center, Program Director and Senior 
	Scientist at Teknowledge, Inc. and Program Manager at Advanced Decision
	Systems.  He has also held a faculty position at Rutgers University.

	As Director of Stanford's Intelligent Systems Laboratory (ISL), Prof. 
	Fehling is continuing to lead interdisciplinary research on "resource-
	constrained problem-solving" by individual problem-solving agents as 
	well as by multiple, interacting agents. Prof. Fehling is investigating
	how intelligent problem-solving agents dynamically adapt their reason-
	ing and decision-making activities to satisfy contextually determined 
	constraints.  More specifically, this research is producing (1) new 
	methods for meeting real-time performance constraints in problem-
	solving applications and (2) practical methods for managing uncertainty.
  	These methods are being used to enhance both centralized and highly 
	distributed forms of problem-solving.  This interdisciplinary research
 	combines computational methods from AI and computer science with 
	quantitative, analytic approaches from the systems sciences.
 
	Prof. Fehling's work contributes significantly to the growth of new, 
	practical technology.  Prof. Fehling and his colleagues and students 
	developed and deployed many prototype and fully operational systems 
	for important applications such as industrial process control, 
	logistics and factory scheduling, and control of advanced aerospace 
	systems.  Prof. Fehling holds a patent for the development of CIM-
	FLEX-Teknowledge's ABE(TM) system, a new-generation software develop-
	ment environment for constructing large-scale, distributed systems 
	that combine AI and non-AI computational methods.  
 
67.70DECvoice Class -- ** Interesting **HERON::ROACHTANSTAAFL !Wed Feb 07 1990 10:4043
Printed by: Pat Roach                                    Document Number: 010288
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     06-Feb-1990 08:19pm CET
                                        From:     EVANS
                                                  EVANS@AITG@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DECvoice Class

From:	POBOX::SZALAJKA     "AI in Chicago"  1-FEB-1990 14:41:09.35
To:	AITG::DOUGLASS
CC:	
Subj:	DECvoice/CA-Expert Class

    For your information (please spread the word) we are offering
a DECvoice/CA-Expert class at the Educational Services facility
at ACI outside Chicago next week, February 5-9, 1990.  The class
is open to customers and internal people.  The course enrollment
is now open beyond our district because spaces remain unfilled.
Enrollment can be made with the Central Area Registrar at
708-806-7578 or DTN 474-7578.  I will also be in my office
all day Friday, Feb. 2nd, at CPO to answer any questions about
the class (DTN 447-2817 or 312-419-2817).  The class is being
taught by Computer Associates.  Mail can also be sent
to the Central Area Registrar at CEA_REGISTRAR @ACI.

    CA-Expert, formerly Application Expert, is a backward-chaining
expert system tool used to build DECvoice applications; it has 
automatic call built into it for the DECvoice sub-routines.

    I also asked the Central Area REgistrar to make John Giger
and SELECT aware of these class openings.  I told JOhn I would
like to use SELECT to build DECvoice applications in the future.

                                   Walter S. Szalajka @CPO
                                   DTN 447-2817
67.71RICH+BARNETT, 2/9,9:30HERON::ROACHTANSTAAFL !Fri Feb 09 1990 11:33121
Printed by: Pat Roach                                    Document Number: 010346
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-Feb-1990 10:54pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: RICH+BARNETT, 2/9,9:30


   




	TITLE:     	"Knowledge-Based Natural Language"

	SPEAKERS:  	JIM BARNETT, KBNL Group Leader, and

		   	DR. ELAINE RICH, Co-director
			Artificial Intelligence Laboratory
	   		Microelectronics and Computer Technology 
			Corporation (MCC)

	PLACE:     	Cafeteria, DLB5

	DATE:      	Friday, February 9, 1990

	TIME:      	9:30-11:00 A.M.

	HOST:      	BILL KUHLMAN
	   		MCC Liaison, Corporate Research




	Natural language and knowledge about the world go together.  This 
	is hardly a new observation.  But it is only beginning to be a  
	correct one from a computational point of view, because, although 
	it is true for the world as a whole and a language as a whole, it 
	is usually not true for tiny slices out of either of them.  Yet 
	the best computational models have been of only relatively tiny 
	slices, particularly of the world, and to a lesser extent, of  
	language.

	In this talk, we will describe our work on a Knowledge Based Natural 
	Language System, which relies on the CYC common sense knowledge base 
	that Doug Lenat is building at MCC.  

	A key idea behind this work is that the natural relationships that 
	exist between the way the world is structured and the way we use 
	language to talk about the world introduce redundancy into the 
	process of creating a knowledge base about the world and a parallel 
	knowledge base about a language.


	This redundancy can be used:  to simplify the task of creating a 
	KB about the world; to simplify the task of defining a lexicon for 
	NL processing; to detect inconsistencies in both the KB and the 
	lexicon; and to produce a resulting NL system that is far less 
	brittle than existing ones.






	"Knowledge-Based Natural Language"
	Jim Barnett and Dr. Elaine Rich
	Page Two



	The KBNL system is composed of five subsystems:

   		1. Lucy, an English understanding system

 	  	2. Koko, an English generation system

   		3. Tell, a knowledge-based lexical acquisition 
		   system that supports both Lucy and Koko

   		4. Show, a tool for retrieving objects from a 
	           knowledge base using English descriptions

   		5. Scan, a full text retrieval system based 
                   on Lucy and Koko.

	The various stages of development of these systems will also be 
	described.

	Jim Barnett is the group leader of the KBNL project at MCC, and has  
	been the major architect of the semantic interpretation component 
	of that system since joining the group several years ago.  

	He has a B.A. in History from Connecticut College, an M.S. in 
	Linguistics from Cornell University, and is currently pursuing a  
	Ph.D. in Linguistics at the Univertity of Texas, where his disser-
	tation work is focusing on disambiguation strategies.

	Elaine Rich is co-director of the Artificial Intelligence Lab in 
	MCC's Advanced Computing Technology (ACT) Program, where she is 
	currently responsible for two projects:  KBNL, a large-scale,  
	knowledge-based natural language processing effort; and RA, whose  
	goal is to produce a platform that supports the construction of  
	hybrid and distributed knowledge-based systems.  

	Prior to joining the AI lab, she was associate director of the  
	Human Interface Laboratory at MCC, where she developed an  
	architecture for intelligent, multimodal interfaces.  She was 
	an assistant professor of computer sciences at the  University  
	of Texas at Austin prior to joining MCC.  She received an A.B.
	in linguistics and applied mathematics from Brown University  
	and a Ph.D. in computer science from Carnegie-Mellon University.
67.72YESHERON::ROACHTANSTAAFL !Fri Feb 16 1990 10:0349
Printed by: Pat Roach                                    Document Number: 010453
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     16-Feb-1990 02:01am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEXPERT OBJECT TM

          	PROGRAMMING IN NEXPERT OBJECT TM TRAINING


          Digital Educational Services offers NEXPERT Object TM
          training on a regular basis at the Digital Training Center
          in Marlborough, Massachusetts.

	  Our next offering of this class is February 26, 1990

	  
          ABOUT THE COURSE

          Your job is to implement solutions for real problems.
          Problems involving process control, real-time diagnostics
          and monitoring in manufacturing, design and configuration
          in electronic and mechanical engineering, or classification
          and intelligent decision support in finance and insurance.
          These are real problems you face every day. With
          instruction based upon real-world experience and hands-on
          use of NEXPERT Object TM, you can gain the skills needed to
          solve real problems with Expert Systems.

          You can ask questions that are most important to your
          business and develop the answers on a Digital VAXstation.

          REGISTRATION, INFORMATION

          For registration call the Educational Services Registration
          office in Bedford, MA:  (617)276-4380.

          For more information on the NEXPERTTM  course, call
          Educational Services at the Digital AI Training Center in
          Marlborough, MA:  (508)490-8893.
67.73Object-Oriented Design in CHERON::ROACHTANSTAAFL !Mon Feb 26 1990 12:4579
Printed by: Pat Roach                                    Document Number: 010531
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Feb-1990 06:04pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Object-Oriented Design in C

                    OBJECT ORIENTED DESIGN IN C


         COURSE NUMBER:     OODC001
         DATE:              MARCH 27-30, 1990
         TIME:              8:30 - 5:00
         LOCATION:          ENGINEERING TRAINING ROOM TAY2-1
         INSTRUCTOR:        JOHN HOLIDAY
         COST:              $1000

------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
WHO SHOULD ATTEND:
C Programmers who need to understand the Object Oriented paradigm but 
do not want/need to learn C++.
This course is not a substitute for an introductory course in "C".
Current experience in C is required.

PREREQUISITE
Programming experience in C.

DESCRIPTION:
This course is a four day introduction to the principles of object
oriented design for the development of application programs in the
C programming language.  All the features of the C programming language 
are organized around the object oriented modelling paradigm to teach
students: a) the benefits of organizing software solutions into an
object oriented framework, and b) how object oriented design offers a
rich modelling tool with tremendous opportunity for software re-use.
Ten "hands on" laboratories develop a practical object-based software 
tool that they can re-use in their individual software applications.
For students who are proficient in C, each lab assignment has an 
advanced segment which gives them an opportunity to experiment with 
more sophisticated aspects of C and object oriented programming.

TOPICS: 	
	o  Introduction to Object Oriented Design
	o  Representation of Object Data Properties
	     - Built-in Types of the C Language
             - User Defined Types: Structs, Unions, Pointers, Arrays
             - Encapsulation of Data into Objects
	o  Representation of Object Behavior
	     - Functions: Expressions, Flow of Control
             - functions as Object Methods
	o  Instantiation and Destruction: Scope
        o  Data Hiding: Static Declarations
	o  Data Abstraction: Factoring Commonality
        o  Object composition: Layering Data Abstractions
	o  Inheritance: Re-Using Communality
	o  Method Abstraction: Message Sending
	o  Design of Object Oriented Frameworks
	o  Preview of Features of C++

OBJECTIVES:  
Upon Completion of this course, students will be able to:
	o  Translate the concepts of object oriented design
           into good C code
	o  Organize software developments into frameworks that
           promote re-use of software
        o  Develop C software that is amenable for re-use in 
           (future) C++ programming environments
67.74PC INTEGRATION, BUSINESS UNIT/PRODUCT INFORMATIONHERON::ROACHTANSTAAFL !Mon Feb 26 1990 12:4763
Printed by: Pat Roach                                    Document Number: 010536
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Feb-1990 07:03pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PC INTEGRATION, BUSINESS UNIT/PRODUCT INFORMATION

Title             PC INTEGRATION, BUSINESS UNIT/PRODUCT INFORMATION SERIES    
Instructor(s)     PCI-PBU Product/Process Managers
Date              22-FEB-90
Time              1:00 -  5:00
Location          Gen. Doriot Auditorium MLO4-5 
Course Number     TSS90-25
Course Fee        $   50
Registration      Registration is REQUIRED...SET HOST MILRAT
                  Username and Password:COURSES
Cancellation      Withdrawal must occur not later than 10 work days prior 
		  to event to avoid being charged.


Agenda            1:00 - 2:15   PCI STRATEGY     VIJAY THAKUR
February 22nd     				 (PCI-PBU MGR)
                                                                                
                  2:15 - 3:00   PCI PRODUCTS
                                                                                
                   	-VMS Services and MS DOS Clients
                    	 ANITA UHLER (Product Mgr)
                                                                                
                   	-PC LAN/Server 3100
                         RON GEMMA (Product Mgr)
                  
                   	-OS/2 Server and Client
                         MARLENE STEGER (Product Mgr)
                                                                                
                  3:00 - 3:15  BREAK
                                                                                
                  3:15 - 3:45 

			-ULTRIX Services
	                 JANE MURPHY (Product Mgr)

                        -MAC Integration
                   	 DAVE GLASSON (Product Mgr)
                                                                                
                  3:45 - 4:15  PCI Marketing Program
                   	 DEANE CURRAN (BPM MGR)
                                                                                
                  4:15 - 4:45  PCI DEMO
                   	 DON CALL (Tech Support Mgr)
                                                                                
                  4:45 - 5:00  Question and Answer
                                                                                
                  NOTE: Details for additions will be provided as
                        they become available.
67.75Cambridge Lab TodayHERON::ROACHTANSTAAFL !Mon Feb 26 1990 14:2970
Printed by: Pat Roach                                    Document Number: 010566
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Feb-1990 06:50pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Cambridge Lab Today

REMINDER: 	       Digital Equipment Corporation
			   Cambridge Research Lab
				  Seminar
  
			 Friday, February 23, 1990
			      3:00pm - 4:00pm
  
	  Bounded Concurrent Time-Stamp Systems Are Constructible
				 Nir Shavit
			     Hebrew University
				 Jerusalem
				   
Concurrent timestamping is at the heart of solutions to some of the most
fundamental problems in distributed computing.  Based on concurrent timestamp
systems, elegant and simple solutions to core problems such as
FCFS-mutual-exclusion, construction of multi-reader-multi-writer atomic
registers, probabilistic consensus, and others were developed.  Though
constructions of bounded timestamp schemes for the sequential case have been
shown by Israeli and Li (1987), the only known implementation of a concurrent
timestamp system is theoretically unsatisfying, since it requires unbounded
size timestamps, in other words, unbounded memory.  Not knowing if bounded
concurrent timestamp systems are at all constructible, researchers were led
to constructing complicated problem-specific solutions to replace the simple
unbounded ones.  In this work, for the first time, a bounded implementation
of a concurrent timestamp system is presented. It provides a modular
unbounded-to-bounded transformation of the simple unbounded solutions to
problems such as above.  It allows solutions to two formerly open problems,
the bounded-probabilistic-consensus problem of Abrahamson (1988) and the
FIFO-$\ell$-exclusion problem of Fischer, Lynch, Burns and Borodin (1979),
and a more efficient construction of MRMW atomic registers.
 
Joint work with Danny Dolev.
-------------------------------------------------------------------------
Nir Shavit received the B.A. and M.Sc. degrees in Computer Science from
the Technion, Israel Institute of Technology in 1984 and 1986,
and will receive the Ph.D.  degree in Computer Science from the Hebrew
University in Jerusalem this year. He will be a visiting scientist at
IBM Almaden research center starting March. His research interests include
concurrency control and synchronization in multiprocessor systems, as well
as various aspects of distributed network algorithms.
-------------------------------------------------------------------------
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and
Broadway.  The address is:
 
                       Digital Equipment Corporation
                       Cambridge Research Lab
                       One Kendall Square, Bldg. 700
                       Cambridge, Mass 02139
 
For questions regarding the seminar, directions, or parking, or to add
your name to the mailing list, call DTN 259-6601 or send email to 
crl::tss.
 
67.76Next EPITOOL OfferingHERON::ROACHTANSTAAFL !Tue Feb 27 1990 10:1581
Printed by: Pat Roach                                    Document Number: 010608
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     26-Feb-1990 07:48pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Next EPITOOL Offering

             ******************************************************
             *                                                    *
             *       Educational Services Intelligent Systems     *
             *                                                    *
	     *		      is happy to announce                *
             *                                                    *
             *            the next offering of EPITOOL            *
             ******************************************************

	
Dates:		March 12-16, 1990

Location:	Artic Classroom - DLB12-1

Price:          $1090 per person

Registration:	To register for this course call the Registration Office
	        at dtn 249-4380.

Information:	For more information contact 
		Malin Goodwin at ESIS::GOODWIN or DTN 291-8894
		
About EPITOOL:
--------------
	EPITOOL is a Lisp based hybrid expert system tool developed by 
	Epitec AB in Link�ping, Sweden. As announced at IJCAI in Detroit,
	EPITOOL will be a software product distributed and sold by Digital.

	EPITOOL	has a wide range of knowledge representations including 
	forward and backward chaining rules, objects and methods, and 
	functions for represenation of procedural knowledge. Epitool also
	provides functions for creation of Decwindows-style application
	graphics, and call-in  call-out to any VAX supported language.

	Development of Knowledge bases can either be done on a VAX station,
	using development graphics, windows and menus, or on a regular
	terminal using Epitools own command language.  
        
About the course:
-----------------
	        
	The course consists of lectures and labs. Labs will provide the 
	students with a hands on experience working with the tool. The
	topics covered in the course are:

	- Knowledge Representation in Epitool: 
		Concepts, Aspects, Individuals, Objects-Methods.

	- Actions:
	  	Rules, Backward and Forward chaining. Functions.
		Rule sets.

	- Development tools:
		Error handling, Debugging

	- Values:
		Uncertain, Unknown, Descriptions, Groups.

	- End user interface:
		Questions, Application graphics, Explanations,
		Streams and active paths

	- Integration

	- Customizing the development environment.
67.77DECdecision Course Offering On-Site - for AITC ONLYHERON::ROACHTANSTAAFL !Tue Mar 06 1990 08:36101
Printed by: Pat Roach                                    Document Number: 010723
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     05-Mar-1990 07:30pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DECdecision Course Offering On-Site - for AITC ONLY






			         "Basic DECdecision"
			         ===================

	1.  COURSE DESCRIPTION:

		This lecture/lab course is designed to teach basic concepts, 
		features and functions of the four DECdecision components.

		Students learn to open and query Access tables, generate
		reports, transfer data across components, create and update
		Calc tables, control the display format of data, draw charts,
		modify chart characteristics and add overlays, and record
		repetitive tasks.
		
	    Objectives:
	    ----------

		o OPEN and query Access tables
		o IDENTIFY and set Access column characteristics
		o CREATE and print Access reports with line and page breaks
		o CREATE Access folders and tables
		o EDIT and maintain Access tables
		o CHART Access data
		o USE object help
		o CREATE Calc tables
		o DEFINE Calc data types and cell characteristics
		o USE relative, mixed and absolute cell references
		o USE Calc functions
		o CHART and "live Link" of Calc data
		o MODIFY chart attributes
		o RECORD repetitive tasks using Building
		o MODIFY recorded tasks using Splice mode

 	    Prerequisites:
	    -------------

		Using DECwindows or the equivalent.  One should be able to:
		
		o Use the mouse
		o Give input focus
		o Shrink windows to icons
		o Move a window
		o Change the size of a window
		o View and select menu items
		o Use scroll bars
		o Select dialog box options
		o Work with the Session Manager and Fileview windows

	    Length:	2 Days
	    ------






	2.  COURSE DATES:

		o May 10-11

	3.  COURSE LOCATION AND TIME:

		DLB12-1, Mediterranean Classroom (Classroom 2) 8:30-5PM.

	4.  COURSE FEE/SIZE:

		Course fee per student is $210.00, and is limited to 12.
		
		The full fee will be charged for those who register, but
		do not attend and provide no alternate.

	5.  REGISTRATION INFORMATION:

		Send the following information to SELECT::HUCKINS:

		o Cost Center
		o Cost Center Manager's Name:
		o Name:
		o NODENAME::NAME:
		o Badge #:
67.78TWO ADVANCED COURSE ANNOUNCEMENTS,HERON::ROACHTANSTAAFL !Thu Mar 15 1990 12:13138
Printed by: Pat Roach                                    Document Number: 010834
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Mar-1990 08:39pm CET
                                        From:     KING
                                                  KING@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: TWO ADVANCED COURSE ANNOUNCEMENTS,


Two advanced AI Training and Education course announcements.  Please send all
replies to AIADM::LEVIN.

********************************************************************************
		       Testing Knowledge-Based Systems
********************************************************************************

DATE:		April 12 - 13, 1990

LOCATION:	Pacific Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA  

DESCRIPTION:    As knowledge-based system technology matures, there are more 
		systems used on a daily basis by a large number and variety
		of users. With this increased impact comes a concern about 
		the reliability of these systems, and a desire to generate 
		more comprehensive testing plans.   This 2-day course describes 
		a variety of methods for testing knowledge-based systems, 
		including validation and verification techniques borrowed 
		from the testing of conventional software, and techniques 
		unique to knowledge-based systems. 


GOALS:  	By the end of this course, the student will:

     		o understand similarities and differences between testing 
           	  knowledge-based systems and testing conventional software;

     		o describe methods of validation and verification over the
        	  project life cycle;

     		o learn a variety of methods for validating and verifying 
		  knowledge-based systems, including regression testing, unit 
		  testing, automated knowledge-base checkers, designing for 
		  testability;

     		o gain exposure to various cases of testing knowledge-based 
		  systems and benefit from lessons learned by engineers 
		  experienced in testing;

     		o learn which testing methods have been successful with which 
        	  languages, tools and types of problems;

     		o describe the testing options in terms of a cost-benefit 
		  metaphor which addresses organizational, technical and 
		  business issues;

     		o be able to generate a high-level testing plan for their own 
        	  application.

INSTRUCTOR:	Theresa Chow, Instructor

AUDIENCE and
PREREQUISITES:	The course is aimed toward engineers or technical managers who 
		are developing knowledge-based systems, and need to develop or 
                implement testing plans for such systems. 

		It is assumed that the student has completed the Expert Systems 
		Training Program or has equivalent knowledge and experience, 
		and that the student has spent at least 6 months involved with 
		knowledge-based system development.   

LENGTH and 
FORMAT:		2 Days - Lecture

COST:		$1000 
    
REGISTER:	Send mail to AIADM::LEVIN.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.





********************************************************************************
	            DESIGNING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

OFFERINGS:	March 20, 1990
		May 29, 1990

LOCATION:	Atlantic Conference Room
		Digital Equipment Corporation
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:  	Knowledge engineers need to produce specifications for the 
		effective development of knowledge-based systems (KBSs).  
		These may be in the form of running prototypes and/or an 
		evolving set of design documents.  The purpose of this course 
		is to provide guidelines for balancing these two alternatives 
		and ensuring that critical design decisions are made regardless 
		of the method used.  The course will focus on the creation of 
		design documents which facilitate the optimal use of proto-
		typing as part of the design process.

GOALS:  	Upon completion of the course, participants will be able to:

		o Recognize the value of design for guiding KBS 
		  development
		o Describe a process which incorporates both planning 
	          and prototyping
		o Identify the most critical design decisions that affect 
	          KBS development
		o Contribute to the creation of a KBS design document

INSTRUCTOR:	Christopher Cherpas

LENGTH: 	1 day

FORMAT: 	Lecture/Discussion

COST:   	$500 

REGISTER:	Send mail to AIADM::LEVIN.  Include badge number, cost 
		center, group and organizational name.
	
67.79ULTRIX FUNDAMENTALS COURSE OFFERINGSHERON::ROACHTANSTAAFL !Tue Mar 20 1990 11:12150
Printed by: Pat Roach                                    Document Number: 010918
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Mar-1990 02:46am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ULTRIX FUNDAMENTALS COURSE OFFERINGS

         COURSE NUMBER:         ULTF006
         DATE:                  MARCH 26-28, 1990
         TIME:                  8:30 - 5:00
         LOCATION:              FREEPORT CLASSROOM LKG2-2/V4
         INSTRUCTOR:            IVAN PREZ-MENDZE
         COST:                  $400
         
         COURSE NUMBER:         ULTF007
         DATE:                  MAY 9-11, 1990
         TIME:                  8:30 - 5:00
         LOCATION:              FREEPORT CLASSROOM LKG2-2/V4
         INSTRUCTOR:            DAVE RUBIO
         COST:                  $400
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
COURSE GOALS:
This course is desgined for the new Ultrix user.  Its goal is to
familiarize the student with the general capabilities of the Ultrix
operating system.  The student will learn to create and modify files
using the text editor vi; manipulate files and directories using
shell (csh) comands; become familiar with file permissions; execute
simple, sequential, and pipeline command lines; and communicate with 
other users.  The student will become familiar with the standard Ultrix
documentation.

KEY TOPICS:
Passwords
File system
File permissions
chmod
Ultrix shell
vi text editor
Regular expressions
Shell meta-characters
I/O redirection
Manual pages 
Pipelines
mail

FORMAT:
Lecture plus Lab exercises

MATERIALS:
Student workbook
Textbook "Introducing the Unix System"

RREQUISITES:
Familiarity with computer system concepts and terminals.


                 COURSE OUTLINE - ULTRIX FUNDAMENTALS

I. INTRODUCTION
   Ultrix features
   Ultrix layering
   Ultrix documentation

II.LOGGIN IN
   Setting your terminal
   Password command
   Ultrix shell
   Ultrix shell
   Ultrix file system

III.COMMUNICATION COMMANDS
   mail
   who
   write
   news

IV.ULTRIX FILE SYSTEM
    File types
    Directories
    Pathnames
    File permissions
    File system commands

V.INTRODUCTION TO VI EDITOR
   Cursor Movement
   Reading/Wtiting Files
   Buffer
   Commands
   Regular expressions
   Searching
   Substituting

VI.ESTABLISHING A LOGIN ENVIRONMENT
   Environment variables
   .cshrc
   .login
   .profile

VII.ULTRIX SHELL
   Features
   Shell meta-charactes
   Command lines

VIII.I/O REDIRECTION AND PIPELINES
   Standard input and output
   Standard error
   Redirection
   Pipelines

IX.FILE SYSTEM COMMANDS
   cp,mc,ln
   ls
   find
   cpio
   tar
   file
   od

X.TEXT PROCESSING COMMANDS
  grep
  sort
  diff
  wc
  pr
  tail
  spell

XI.BACKGROUND COMMAND LINES
  Process ids
  ps
  wait
  kill
  nohup
  Job Control

XII.OVERVIEW OF ULTRIX ENVIRONMENT
  Development tools (compiling C programs)
  Text Processing tools
67.80NTU SATELLITE OFFERING - "Engineering Your Future..HERON::ROACHTANSTAAFL !Tue Mar 20 1990 11:13182
Printed by: Pat Roach                                    Document Number: 010920
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Mar-1990 03:17am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NTU SATELLITE OFFERING - "Engineering Your Future.."

		THE 1990s: ENGINEERING YOUR FUTURE --
			   FOR EXPERIENCED PROFESSIONALS

     COURSE NUMBER:		PDAPR01
     DATE: 			Mondays April 2, 9, 16, 23
				(NOTE: April 16 is a DEC holiday. This
				session will be videotaped and shown at the
				same time, 11AM-1PM, on April 17 in the 
				Boothbay Conference Room.)
     TIME:			11 AM - 1 PM
     LOCATION:			Boothbay Conf. Rm, LKG2-2
     INSTRUCTOR:		Peggy Hutcheson 
     COST:			$250 (includes all four sessions)
--------------------------------------------------------------------------------
     TO REGISTER:  Set Host DELNI; Username: COURSES; Password: not required
                   Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES

**NOTE:	This is the fourth in a six month series which addresses the topic
"The 1990s:  The Future of Engineering and Engineering Your Future."    
The June series will be for technical managers at any level in the 
organization.  In May the development of more effective presentation skills 
will be addressed.  (See separate descriptions for the May and June 
sessions.)
                                                                               
COURSE DESCRIPTION:  

Participants will assess how the changes outlined by the top engineering 
executives in the January series will affect his or her professional 
effectiveness and career direction.  Using workbook materials, each 
individual will identify the most effective ways to tap his or her 
energies and talents as they relate to career issues.  Peggy Hutcheson will 
use experts in the field and organizational case studies to present the 
programs.

(See individual session descriptions below.)

SPEAKER:  

Peggy Hutcheson is a partner in Atlanta Resource Associates, a firm 
dedicated to helping organizations capitalize on the talents and energies 
of their human resources.  She and her firm have worked with a number of 
leading firms, including Coca Cola Co., Digital, IBM, NCR and others, on 
career issues.  She has a Ph.D. from Georgia State University where her 
studies focused on career development as an aspect of organizational 
behavior.  Her undergraduate work was in journalism and mass communications.  
She is a former board member of the American Society for Training and 
Development.

INTENDED AUDIENCE:  

This series is for experienced professionals who need to test career 
directions and/or explore new directions more appropriate to their 
technical environment and areas of interest.

SPONSOR: 

National Technological University
---------------------------------------------------------------------------

	April 2		A BLUEPRINT FOR CAREER SUCCESS:
			EXAMINING SKILLS, INTERESTS AND VALUES


SESSION DESCRIPTION:

Maintaining energy and excitement in your career is an ongoing challenge.  
Technical professionals are most motivated when work stretches your talents 
in ways consistent with some core values.

This program explores your interests, your skill strengths, and your work 
values.  You will examine past successes to develop themes for future 
planning, and you will develop a framework for making decisions about your 
current and future work options.

SESSION OBJECTIVES:  

o Identify the common denominators for successful career development 
  programs.

o Chart your current most valued skills.

o Explore work-related interests.

o Examine how changes in jobs and/or opportunities affect skills and values.

o Develop a profile for career decision-making.

		******************************************

     April 9	MATCHING SKILLS WITH OPPORTUNITIES:  
		ALTERNATIVES FOR GROWING WITHOUT CHANGING YOUR CAREER


SESSION DESCRIPTION: 

Matching individual needs and goals with work options calls for a thorough 
investigation of job characteristics.  As you explore the payoffs of 
different career choices, it becomes evident that "up is not the only way."

This session will provide information to help sort the options available 
and to increase the rewards from your current job.  You will also explore 
key skills needed for success in each stage of a technical career.

SESSION OBJECTIVES:  

o Prioritize important characteristics of work options.

o Identify the strengths and liabilities of different career paths.

o Examine opportunities for growth within current positions.

o Recognize the skills needed to grow in each stage of a technical career.

o Plot possible next career steps.

		********************************************

	April 16  CHARTING YOUR COURSE: MAKING DECISIONS 
		  FOR CAREER GROWTH AND SATISFACTION


SESSION DESCRIPTION:  

Career goals and preferences change over time.  This program provides 
information and assessments to pinpoint the primary career issues for 
technical people in each life/career stage.  You will see how some career 
choices can get you off track from a real sense of success and fulfillment. 
Learning how to apply strategies for making career decisions -- large and 
small -- will move you toward important personal goals.

SESSION OBJECTIVES:  

o Examine the sequence and challenges of life/career issues.

o Explore ways to enrich decision-making processes.

o Use techniques to make more satisfying career decisions.

o Develop goal statements consistent with personal career profiles.

		********************************************

	April 23	OVERCOMING OBSTACLES:
			TOUGH ISSUES FOR TECHNICAL CAREERS


SESSION DESCRIPTION:  

Changing organizational structures, markets and technologies redefine what 
it takes to plan and manage careers.  This program guides you through a 
process to identify the hurdles that may appear to block your career.  You 
will learn skills and strategies for keeping your career on track, 
regardless of the obstacles.  Developing a specific action plan for 
managing your toughest career issues is a critical component for taking 
charge of your work life.  

SESSION OBJECTIVES:  

o Identify realities within organizations and the marketplace that impact 
  technical careers.

o Examine skills and strategies for accomplishing goals within these 
  organizational realities.

o Learn specific skills to better manage the work environment and maintain 
  career momentum.
67.81CRL - Exp. Sys. Applied to Graphical User InterfaceHERON::ROACHTANSTAAFL !Tue Mar 20 1990 11:1498
Printed by: Pat Roach                                    Document Number: 010921
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Mar-1990 03:32am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: CRL - Exp. Sys. Applied to Graphical User Interface Design

		    Digital Equipment Corporation
			Cambridge Research Lab
			       Seminar
 
		      Thursday, April 19, 1990
			   3:00pm - 4:00pm
 
      EXPERT SYSTEMS APPLIED TO THE DESIGN OF GRAPHICAL USER INTERFACES
 
                            James D. Foley
 
                            Dept. of EE&CS
                   The George Washington University
                         Washington, DC 20052
                             202-994-4952
                         [email protected]
 
The User Interface Design Environment (UIDE) is an integrated set of tools 
which supports the designers of graphical user interfaces.  A knowledge-base 
representation of the interface is created by the designer, and is used in 
turn to help the designer refine and improve the design, and to provide 
run-time assistance to the end-user.  The talk overviews UIDE, and includes 
a videotape of animated graphical help generated.using UIDE concepts.  Brief 
video clips of other graphical interface projects at GW will be shown.
 
A user interface design is specified to UIDE in terms of objects, actions, 
attributes, and pre- and post-conditions associated with the actions.  The 
user interface design is represented as a knowledge base.  The design can be:
 
- used to analyze the interface's speed of use
 
- used to generate an operating user interface via SUIMS, our Simple User 
  Interface Management System
 
- used to generate context-sensitive help messages and animated explanations 
  of how to complete tasks
 
- automatically transformed into a different but functionally equivalent 
  interface by application of design paradigms such as currently-selected 
  object, global attribute value, command mode, and specialized commands.  
  These capabilities allow us to automatically generate a series of 
  functionally equivalent user interfaces for evaluation and use
 
- used to create dialogue boxes whose visual structure explicitly reinforces 
  the underlying logical structure of information in the dialogue.
 
--------------------------------------------------------------------------
 
James Foley is Professor and Chairman of Electrical Engineering and
Computer Science at the George Washington University.  He earned his
PhD at the University of Michigan.  His interests include user
interfaces and interactive computer graphics; his research, sponsored
by NSF, NASA, and Siemens, focuses on building UIDE, the User
Interface Design Environment.  He is co-author, with A. van Dam, of
Fundamentals of Interactive Computer Graphics, and is co-author, with
van Dam, S. Feiner, and J. Hughes, of Computer Graphics: Principles
and Practice, which will be published this spring.  Foley is a Fellow
of the IEEE, serves on several editorial boards, and consults
regularly for governmental and industrial organizations.
 
-------------------------------------------------------------------------
 
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and
Broadway.  The address is:
 
                       Digital Equipment Corporation
                       Cambridge Research Lab
                       One Kendall Square, Bldg. 700
                       Cambridge, Mass 02139
 
For questions regarding the seminar, directions, or parking, or to add
your name to the mailing list, call DTN 259-6600 or send email to 
crl::tss.
 
 
% ==== Internet headers and postmarks (see DECWRL::GATEWAY.DOC)
% Received: by easynet.crl.dec.com; id AA16560; Mon, 19 Mar 90 11:21:06 -0500
% Received: by crltrx.crl.dec.com; id AA09985; Mon, 19 Mar 90 11:22:10 -0500
% Message-Id: <[email protected]>
% Date: Mon, 19 Mar 90 11:22:08 EST
% From: maryg
67.82INTEGRATING KB SYSTEMSHERON::ROACHTANSTAAFL !Tue Mar 20 1990 11:1585
Printed by: Pat Roach                                    Document Number: 010923
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Mar-1990 04:17am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: INTEGRATING KB SYSTEMS

	        Please send all replies to AIADM::LEVIN

********************************************************************************
	           INTEGRATING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

       DATE:	April 23 - 25th, 1990

   LOCATION:	Atlantic Conference Room, DLB12-1 
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:    The successful delivery of an application that uses artificial
                intelligence (AI) tools and techniques frequently depends
                on the ability of the system to coexist with other
                applications and existing data.  Using examples and
                exercises from case studies, this course presents
                guidelines and methods for integrating knowledge-based or
                expert systems into the application environment and easing
                a system's transition from prototype to production.  The
                course will highlight issues in project planning, system
                life cycle, knowledge representation, and implementation
                that must be addressed to successfully integrate a
                knowledge-based system into a business application
                portfolio.

      GOALS:    Upon completion of the course, participants will be able
                to:

                    - Define integration requirements for knowledge-based
                      applications

                    - Incorporate integration issues into the project plan

                    - Define the relationship (and a transformation)
                      between a relational data definition and a
                      knowledge representation

                    - Write simple VAX LISP programs that access non-LISP
                      routines and/or data files

                    - Identify trends toward the integration of databases
                      and AI applications
 
 PRESENTER:	John Arnold

  AUDIENCE:     Application developers, programmers, or software engineers
                involved with the delivery of knowledge-based or other AI
                systems. 
                Project leaders or technical managers of applications that
                have or are considering a knowledge-based component.


LENGTH & FORMAT:    3 days, Lecture/Lab


PREREQUISITES:  - Completion of the Expert Systems Training Program or
                  equivalent

                - Applications development experience

                - Familiarity with LISP programming

        COST:   $1,500.

    REGISTER:	Send mail to AIADM::LEVIN.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.83OVERVIEW OF AI COURSEHERON::ROACHTANSTAAFL !Fri Mar 23 1990 10:1068
Printed by: Pat Roach                                    Document Number: 011001
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Mar-1990 01:23am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: OVERVIEW OF AI COURSE

Send all replies to AIADM::LEVIN.  Registration requires badge number, cost
center, manager's name and organizational name.

********************************************************************************
                OVERVIEW OF AI/KNOWLEDGE-BASED SYSTEMS
*******************************************************************************

LOCATION:       DLB12-1, Pacific Conference Room
	        295 Donald Lynch Boulevard
	        Marlborough, MA  

DATE:		April 30, 1990

DESCRIPTION:	This is a one-day overview of the fields of Artificial 
		Intelligence (AI) and Knowledge-based systems (KBS).  It 
		is presented by MIT Professors Patrick Winston and Randall 
		Davis, both world renowned leaders in the field.  The first 
		presentation, "Foundations of AI", addresses the definition 
		of AI, its major sub-areas, what it can do, the nature and 
		use of shells and the future of the technology.

		The second presentation focuses on "Knowledge-based Systems."
  		Topics presented include: definitions of these systems, 
		different architectures, and finding and evaluating appropriate 
		applications. In addition, the impacts of technical, business 
		and organizational issues on implementing knowledge-based 
		systems are discussed.

GOALS: 		Participants will be able to:

		o Describe the historical context for the development of 
	  	  AI technology and the key milestones.
		o Identify the different technologies that comprise the field 
	  	  of AI.
		o Understand some of the future directions of AI technology.
		o Describe some of the different architectures used in 
	  	  knowledge-based systems.
		o Discuss some of the issues involved in selecting an 
		  appropriate application for knowledge-based systems 
		  technology.

LENGTH: 	1 day

FORMAT: 	Lecture

PRESENTERS:	Professors Randall Davis and Patrick Winston, MIT

COST: 		$500.

NOTE: 		Although this 1 day course is offered as an entity, it also
		serves as the first day of the INTRODUCTION TO AI IN BUSINESS 
		course.
67.84AITC IS SPONSORING A DAY OF TECHNICAL EXCHANGE WITHHERON::ROACHTANSTAAFL !Mon Mar 26 1990 13:02102
Printed by: Pat Roach                                    Document Number: 011041
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Mar-1990 10:23pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AITC IS SPONSORING A DAY OF TECHNICAL EXCHANGE WITH NIPPON ELECTRIC

AS PART OF A TECHNICAL EXCHANGE BETWEEN NEC (NIPPON ELECTRIC CORP.) AND THE 
AI TECHNOLOGY CENTER, THE FOLLOWING TSS' WILL BE PRESENTED BY NEC:


-->	DATE:		Friday, March 30, 1990

-->	PLACE:		PACIFIC ROOM, DLB12-1

-->	AGENDA:		TSS Schedule as Follows:


	   TIME:               TOPIC:                           SPEAKER:

	9:00-9:10AM	Introductions 			       MITCH TSENG

	9:10-9:50AM	"Image Processing Technology for       NOBUYUKI NARUO
		        Automatic Visual Inspection"           of NEC
      
	9:50-10:30AM	"Concept and Plan for Applying AI      TAKEO NOZAKI
	                Technology in Production Engineering   of NEC
			Area" (NEC)	

++***************************************************************************++

			ABSTRACT 1ST PRESENTATION

	IMAGE PROCESSING TECHNOLOGY FOR AUTOMATIC VISUAL INSPECTION (30min)
			Nobuyuki Naruo  Supervisor
		Production Engineering Development Laboratory
				
		1. Image Processing Technology in Production Engineering
				Developed Inspection Systems
		2. Wafer Inspection System
		3. Solder Joint Inspection System
		4. Future Plan
			(Discussion (10 min)


		Nobuyuki Naruo (Supervisor) BIO

	He received the B.E. degree in Physical Electronics and M.E. degree
	in Applied Electronics from Tokyo Institute of Technology in 1980
	and 1982 respectively.  He joined NEC Corporation in 1982 and has 
	been a member of Production Engineering Development Laboratory ever
	since.  He was a visiting researcher at Purdue University from
	1988 to 1989.  And he is a member of Information Processing Society 
	of Japan.
==============================================================================
			ABSTRACT 2ND PRESENTATION

	CONCEPT AND PLAN FOR APPLYING AI TECHNOLOGY IN PRODUCTION ENG.
			AREA IN NEC (30min)

		Takeo Nozaki, Engineer, Production Engineering Develop-
			ment Laboratory
		Naoko Shimazu, Engineer, Transmission Division
		Nobuyuki Naruo, Supervisor
	
		1. Introduction, Problem Domains in Production Engineering
		2. Developed Systems applying AI Technology
		   Expert System for Supporting Adjustment of Analog PCB
		   Troubleshooting Expert System for Multi-Layer Substrate
		   Inspection System
		   Wave Form Recognition by Neural Network Technology
		3. Developing Knowledge Base System
		   Automatic Adjustment System for Microwave Filter
		   Adjustment Support System for Super-high-frequency FET
		4. Future Plan   (Dicussion 10min)
		

		Takeo Nozaki  Engineer  BIO
	
	He received the B.E. degree in Precision Engineering and M.E.
	degree in Information Engineering from Hokkaido University, in
	1960 and 1962 Respectively.  He joined NEC Corporation in 1987
	and has been a member of Production Engineering Development 
	Laboratory.  And he is a member of Society of Instrument and
	Control Engineers.

		
		Naoko Shimazu  Engineer  BIO

	She received the B.E. degree in Optics Engineering from Tokai
	University in 1983.  She joined NEC Corporation in 1983 and has
	been a member of Production Engineering Department in Transmission
	Division.
67.85An Advanced course announcement from AITEHERON::ROACHTANSTAAFL !Tue Mar 27 1990 12:1883
Printed by: Pat Roach                                    Document Number: 011079
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     26-Mar-1990 07:39pm CET
                                        From:     KING
                                                  KING@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: An Advanced course announcement from AITE

********************************************************************************
	           INTEGRATING KNOWLEDGE-BASED SYSTEMS
********************************************************************************

       DATE:	April 23 - 25th, 1990

   LOCATION:	Atlantic Conference Room, DLB12-1 
		295 Donald Lynch Boulevard
		Marlborough, MA

DESCRIPTION:    The successful delivery of an application that uses artificial
                intelligence (AI) tools and techniques frequently depends
                on the ability of the system to coexist with other
                applications and existing data.  Using examples and
                exercises from case studies, this course presents
                guidelines and methods for integrating knowledge-based or
                expert systems into the application environment and easing
                a system's transition from prototype to production.  The
                course will highlight issues in project planning, system
                life cycle, knowledge representation, and implementation
                that must be addressed to successfully integrate a
                knowledge-based system into a business application
                portfolio.

      GOALS:    Upon completion of the course, participants will be able
                to:

                    - Define integration requirements for knowledge-based
                      applications

                    - Incorporate integration issues into the project plan

                    - Define the relationship (and a transformation)
                      between a relational data definition and a
                      knowledge representation

                    - Write simple VAX LISP programs that access non-LISP
                      routines and/or data files

                    - Identify trends toward the integration of databases
                      and AI applications
 
 PRESENTER:	John Arnold

  AUDIENCE:     Application developers, programmers, or software engineers
                involved with the delivery of knowledge-based or other AI
                systems. 
                Project leaders or technical managers of applications that
                have or are considering a knowledge-based component.


LENGTH & FORMAT:    3 days, Lecture/Lab


PREREQUISITES:  - Completion of the Expert Systems Training Program or
                  equivalent

                - Applications development experience

                - Familiarity with LISP programming

        COST:   $1,500.

    REGISTER:	Send mail to AIADM::LEVIN.  Include badge number, cost 
		center, group and organizational name.

CANCELLATION:   10 days prior to the start date of the course.
67.86Two questions to be answered, please.HERON::ROACHTANSTAAFL !Thu Mar 29 1990 00:5342
Printed by: Pat Roach                                    Document Number: 011130
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     28-Mar-1990 09:52pm CET
                                        From:     KING
                                                  KING@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Two questions to be answered, please.


     AITE would like to know what the market is for 2 to Carnegie Group, Inc. 
     (CGI) courses.  Please take a few minutes to answer the following 
     questions:

1.   WOULD YOU BE INTERESTED IN A 7-DAY KNOWLEDGE CRAFT COURSE?

     This course would consist of a 2-day Knowledge Craft (KC) Hands-On 
     Introductory Seminar and a 1-week KC course.  The objective is to give 
     enough hands-on-experience with KC to allow the student to make an inform-
     ed decision about using KC for a knowledge-based application. The student 
     would develop an in depth understanding of CRL, CRL-OPS, and RPM, as well 
     as a thorough understanding of CRL-Prolog, the Command/Window System, the 
     Graphics System and Advanced CRL.
            
2.   INTERESTED IN A 2-DAY PAKS (Simulation and Related Modules) SEMINAR?

     The objective of a 2-day PAK seminar is to provide students with an under-
     standing of how to use the "pak" modules of Knowledge Craft:  OOPpak, 
     Statpak, Simpak, Graphpak and Timepak.  Upon completion, the student will 
     know how to incorporate the functionality of the "pak" modules into an 
     application.  The course is aimed at those students who are able to create 
     and manipulate schemata, make use of relations, create and use canvases, 
     windows and viewports, and create and manipulate graphic objects.
     
 
67.87"Managing Technology" - Series of NTU Satellite OffHERON::ROACHTANSTAAFL !Tue Apr 03 1990 15:23186
Printed by: Pat Roach                                    Document Number: 011204
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     02-Apr-1990 07:14pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: "Managing Technology" - Series of NTU Satellite Offerings

 		MANAGING TECHNOLOGY IN THE GLOBAL MARKETPLACE


     COURSE NUMBER:     MTG0001, MTG0002, MTG0003, MTG0004
     DATE:    		April 26, May 3, 10, 17
     TIME:              11 AM - 5 PM  
     LOCATION:          See individual session descriptions
     SPEAKERS:		See below
     COST:              $350 per day (register individually)
--------------------------------------------------------------------------------
     TO REGISTER:  Set Host DELNI; Username: COURSES; Password: not required
                   Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
--------------------------------------------------------------------------------
This is a four day symposium which focuses on how to manage technology, 
both nationally and internationally.  Each day is listed separately 
in COURSES and must be registered for individually.  The earlier sessions 
are NOT pre-requisites for the later ones.

COURSE OBJECTIVES:

One of the challenges confronting business and government today is 
management of technology, on a national and international scale.  The 
ability to use and capitalize on the fruits of technology requires new 
areas of technical and management expertise.  Most of this growing body of 
knowledge is emanating from industrial and business sectors.  To meet the 
challenge, the academic community is also developing new education and 
research programs as part of the national effort in this high priority 
area.  

The challenge the U.S. faces, whether in industry, government, or business, 
is how to move a product or service from research into development into 
manufacturing into the marketplace in a cost-effective fashion.  NTU is
bringing together experts from higher education, industry, business and 
government for four days of state-of-the-art forums on how to manage 
technology, nationally and internationally.

INTENDED AUDIENCE:

This special symposium is aimed at engineers, scientists and managers at 
all levels.  

SPONSOR:  National Technological University
-----------------------------------------------------------------------

DAY ONE - APRIL 26 - EDGARTOWN CONFERENCE ROOM - LKG2-1/9U

COURSE NUMBER:		MTG0001

COURSE DESCRIPTION:

Today's presentations will consist of the following:

Keynote Address		ECONOMIC GROWTH AND THE COMMERCIALIZATION OF
			TECHNOLOGY IN THE GLOBAL MARKETPLACE

			Ralph Landau, Consulting Professor of Economics and 
			Chemical Engineering, Stanford University, and Vice 
			President, National Academy of Engineering

Roundtable Discussion	MANAGEMENT OF TECHNOLOGY - THE HIDDEN 
			COMPETITIVE ADVANTAGE

			- Peter Likins, President, Lehigh University
			- Alden Bean, Professor of Management and Technology,
			  College of Business and Economics, Lehigh University
			- Herbert I. Fusfeld, Director, Center for Science 
			  and Technology Policy, School of Management, 
			  Rensselaer Polytechnic Institute
			- Dr. Richie Herink, Program Manager, IBM

Address			MANAGING TECHNOLOGY

			Lowell W. Steele, Consultant and author of the 
			recent book, "Managing Technology: The Strategic View."
-------------------------------------------------------------------------

DAY TWO - MAY 3 - FALMOUTH CONFERENCE ROOM - LKG2-1/9W

COURSE NUMBER:		MTG0002

COURSE DESCRIPTION:

Today's presentations will consist of the following:

Address			THE CORPORATE CHALLENGE IN TECHNOLOGY MANAGEMENT

			- William J. Spencer, Group Vice President, Xerox 
			  Corporation
			- John A. Armstrong, Vice President, Science and 
			  Technology, IBM

Address			TECHNOLOGY INNOVATION

			- Richard N. Foster, Director, McKinsey & Company, 
		          Inc., and author of the book, "Innovation: 
			  The Attacker's Advantage"

Address			NEW APPROACHES TO TECHNOLOGY-BASED 
			BUSINESS DEVELOPMENT

			- Edward B. Roberts, Professor of Management of 
			  Technology, Massachusetts Institute of Technology

--------------------------------------------------------------------------

DAY THREE - MAY 10 - EDGARTOWN CONFERENCE ROOM - LKG2-1/9U

COURSE NUMBER:		MTG0003

COURSE DESCRIPTION:

Today's presentations will consist of the following:

Address			TECHNOLOGY AND INTERNATIONAL COMPETITIVENESS

			- Graham R. Mitchell, Director of Planning, 
			  GTE Laboratories, Inc.

Panel			R & D COOPERATION -- U.S. AND JAPAN

			- Robert S. Cutler, Program Evaluation Staff,
		  	  National Science Foundation
			- Fumio Kodama, Director of Research, Japan's 
			  National Institute of Science and Technology 
			  Policy (STA)
			- Lamont Eltinge, Society of Automotive Engineers 
			  (SAE) Executive Fellow, Office of Science and 
			  Technology Policy
			- Robert G. Latorre, Professor and Chairman, 
			  Department of Naval Architecture & Marine 
			  Engineering, University of New Orleans
			- Hisashi Kobayashi, Dean, School of Engineering 
			  and Applied Science, Princeton University

Address			THE MANAGEMENT CHALLENGE IN THE SOVIET UNION

			- Stephen H. Rhinesmith, President, Rhinesmith and 
			  Associates Inc., and President Reagan's Coordinator 
			  for U.S. - Soviet Exchanges
--------------------------------------------------------------------------

DAY FOUR - MAY 17 - EDGARTOWN CONFERENCE ROOM - LKG2-1/9U

COURSE NUMBER:		MTG0004

COURSE DESCRIPTION:

Today's presentations will consist of the following:

Address			TRENDS IN SEMICONDUCTOR TECHNOLOGY 
			AND THE CHALLENGE

			- Larry W. Sumney, President, Semiconductor 
			  Research Corporation (SRC)

Address			MANAGING DUAL-USE TECHNOLOGIES

			- Lewis M. Branscomb, Albert Pratt Public Service 
			  Professor, John F. Kennedy School of Government, 
			  Harvard University

Address			TECHNOLOGICAL FOUNDATIONS OF STRATEGIC MANAGEMENT

			- Paul S. Adler, Assistant Professor of Engineering 
			  Management, School of Engineering, 
			  Stanford University

--------------------------------------------------------------------------
For a more detailed description of the presentations, please contact Betsy 
Mills, DELNI::MILLS,  or 226-7204.
67.88Need to "sell" your ideas or influence the decisionHERON::ROACHTANSTAAFL !Tue Apr 03 1990 15:2474
Printed by: Pat Roach                                    Document Number: 011205
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     02-Apr-1990 07:14pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Need to "sell" your ideas or influence the decision makers???  Try......"Positive Power and Influence", 6/4, 8:30-5

                 POSITIVE POWER AND INFLUENCE


         COURSE NUMBER:       PPI0006
         DATE:                May 1 - 4, 1990
         TIME:                8:30 - 5:00
         LOCATION:            Pacific c/r DLB12-1
         INSTRUCTOR:          Marilyn Ullman/Digital Management Education
         COST:                $850
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
AUDIENCE:

For Supervisors and Managers. (Individual contributors, team leaders and
project leaders should enroll in "Influence Skills for Technical 
Professionals".) 


COURSE DESCRIPTION

The four-day Positive Power and Influence Workshop heightens
awareness of the personal power participants exert regardless of
position or authority.  Participants learn how to achieve a high
level of personal effectiveness -- get things done through others,
have their ideas adopted and motivate people.

Before the Workshop participants will be provided feedback forms to
collect information about their styles of influence from colleagues.
During the Workshop, participants will examine five basic styles
of influence and twelve behaviors.

Course Objectives:

  Participants will be able to:

	o  Describe five basic influence styles
        o  Describe twelve behaviors
	o  Diagnose influencing situations and formulate effective
           influence strategies
	o  Practice a variety of influencing behaviors
	o  Demonstrate flexibility in influencing others
	o  Influence others without resorting to line authority

Content:

	I  Five basic influence styles 
       II  Twelve Influence behaviors
      III  Using personal power constructively
       IV  Diagnosing influence situations
        V  Developing influence strategies
       VI  Application planning

Methodology:

Large and small group exercises, video-taped role-plays, Influence 
Style Questionnaires, lecturettes and application planning.
67.89NEXPERT OBJECT TrainingHERON::ROACHTANSTAAFL !Tue Apr 03 1990 15:2655
Printed by: Pat Roach                                    Document Number: 011209
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     02-Apr-1990 08:30pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEXPERT OBJECT Training

	**************************************************************
	*   If you receive more than one copy of this message, and   *
	*   DO NOT have multiple accounts, please reply to AIADM::TSS*
	*   right away.  However, please send course registration    *
  	*   information as directed below.  Thank you.               *
        **************************************************************

          PROGRAMMING IN NEXPERT OBJECT TRAINING

          Digital Educational Services offers NEXPERT Object 
          training on a regular basis at the Digital Training Center
          in Marlborough, Massachusetts.

	  Our next offering of this class is April 23, 1990, this
	  is a five day course.  

          ABOUT THE COURSE

          Your job is to implement solutions for real problems.
          Problems involving process control, real-time diagnostics
          and monitoring in manufacturing, design and configuration
          in electronic and mechanical engineering, or classification
          and intelligent decision support in finance and insurance.
          These are real problems you face every day. With
          instruction based upon real-world experience and hands-on
          use of NEXPERT Object, you can gain the skills needed to
          solve real problems with Expert Systems.

          You can ask questions that are most important to your
          business and develop the answers on a Digital VAXstation.

          REGISTRATION, INFORMATION

          For registration call the Educational Services Registration
          office in Bedford, MA:  (617)276-4380.

          For more information on the NEXPERT course, call
          Educational Services at the Digital AI Training Center in
          Marlborough, MA:  (508)490-889
67.90WRITING FOR RESULTSHERON::ROACHTANSTAAFL !Tue Apr 03 1990 15:2750
Printed by: Pat Roach                                    Document Number: 011210
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     02-Apr-1990 10:16pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: WRITING FOR RESULTS

         COURSE NUMBER:   WRIT001
         DATE:            APRIL 19,20, 1990
         TIME:            8:30 - 5:00
         LOCATION:        DLB5-3  Leopard c/r Donald Lynch Blvd.
         INSTRUCTOR:      Linda Marks
         COST:            $600
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
Course Name:	Writing for Results
Length:		2 days

Intended Audience:

Employees wishing to increase the effectiveness of their memos and reports

Course Description:

This workshop presents both the theory and practice of clear, concise, and
effective writing.  Participants will learn how to increase the impact of
their writing and reduce costly miscommunication.  Concepts covered include
dealing with writer's block, organization, work choice, and editing for
impact and audience considerations.  Participants will have an opportunity
to apply guidelines discussed in the workshop to both written exercises and
on-the-job writing tasks.

Course Objectives:

Participants will:

o  Identify a procedure for writing memos and reports
o  Assess strengths and weaknesses in personal writing style 
o  Identify ways to increase the effectiveness of written communication
67.91EPITOOL OFFERINGHERON::ROACHTANSTAAFL !Wed Apr 18 1990 13:1678
Printed by: Pat Roach                                    Document Number: 011355
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Apr-1990 11:02pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL OFFERING

             ******************************************************
             *                                                    *
             *       Educational Services Intelligent Systems     *
             *                                                    *
	     *		      is happy to announce                *
             *                                                    *
             *            the next offering of EPITOOL            *
             ******************************************************
	
Dates:		April 30-May 4, 1990

Location:	Caribbean Classroom - DLB12-1

Registration:	To register for this course call the Registration Office
	        at dtn 249-4380.

Information:	For more information contact 
		Malin Goodwin at ESIS::GOODWIN or DTN 291-8894

About EPITOOL:
--------------
	EPITOOL is a Lisp based hybrid expert system tool developed by 
	Epitec AB in Link�ping, Sweden. As announced at IJCAI in Detroit,
	EPITOOL will be a software product distributed and sold by Digital.

	EPITOOL	has a wide range of knowledge representations including 
	forward and backward chaining rules, objects and methods, and 
	functions for represenation of procedural knowledge. Epitool also
	provides functions for creation of Decwindows-style application
	graphics, and call-in  call-out to any VAX supported language.

	Development of Knowledge bases can either be done on a VAX station,
	using development graphics, windows and menus, or on a regular
	terminal using Epitools own command language.  
	
About the course:
-----------------
	        
	The course consists of lectures and labs. Labs will provide the 
	students with a hands on experience working with the tool. The
	topics covered in the course are:

	- Knowledge Representation in Epitool: 
		Concepts, Aspects, Individuals, Objects-Methods.

	- Actions:
	  	Rules, Backward and Forward chaining. Functions.
		Rule sets.

	- Development tools:
		Error handling, Debugging

	- Values:
		Uncertain, Unknown, Descriptions, Groups.

	- End user interface:
		Questions, Application graphics, Explanations,
		Streams and active paths

	- Integration

	- Customizing the development environment.
67.92Prof. Agogino-4/20-9:30AMHERON::ROACHTANSTAAFL !Wed Apr 18 1990 15:31116
Printed by: Pat Roach                                    Document Number: 011384
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     18-Apr-1990 10:05am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Prof. Agogino-4/20-9:30AM



	TITLE:		"THE CONCEPT DATABASE:  Systematic Concept 
			Search, Design Evaluation, and Archival for 
			Concurrent Design"

	SPEAKER:	Alice M. Agogino, Associate Professor
			University of California at Berkeley

	DATE:		Friday, April 20, 1990

	TIME:		9:30-11:00AM

	PLACE:		DLB5 Cafeteria

	HOST:		Frank Lynch, Manager
			Advanced Systems and Tools Group
			AI Technology Center

	CONCURRENT DESIGN is an approach to design in which simultaneous
	consideration is given to all of the factors involved in the life
	cycle of a product.  Subsets of this idea are called "design for
	*ability", where "*bility" is replaced by manufacturability, main-
	tainability, marketability, and so so.

	Most computer-aided design tools are used primarily during the 
	later stages of the design process for documentation or evaluation 
	of completed designs.  They capture details of design "form" at
	best, and leave out essential aspects of the history of the design
	process:  design intent, intermediate goals, assumptions, feature
	justification, computer simulations and analyses.  

	Further, current design tools are not integrated with the other 
	information systems of the enterprise, and provide little assistance 
	in capturing or transmitting the information resources of the 
	enterprise used to develop designs.  For example, product drawings 
	give very little information for downstream designers to evaluate 
	the impact of a design change.  Because of its importance in life 
	cycle design costs, our focus in this research is at the conceptual 
	stage of design.

	Recent studies have shown life cycle costs of a product are 
	primarily determined during the early, conceptual stages of the 
		design process.  

	The Institute for Defense Analysis reports that 80-90% of U.S. 
	aerospace downstream costs (including fabrication, manufacturing 
	construction, energy, maintenance and disposal) are determined 
	in the first 10% of the design process.  Similar results are 
	reported from the automotive, computer and consumer products 
	industries.  





	"THE CONCEPT DATABASE..."
	Friday, 4/20, 9:30AM
	Prof. Alice Agogino
	Page Two



	Thus the maximum leverage on cost reduction and performance 
	enhancement is during the early stage of design.  Yet most tech-
	niques for "design for manufacturability" are now only applied 
	after product design is near complete and only detailing changes 
	are possible.

	We propose to develop a novel computer-aided concurrent design
	tool for assisting engineers during the conceptual phase of 
	design.  

	The Concept Database will utilize database and optical disk 
	technology to provide the designer with access to a broad 
	spectrum of design information which will suggest, describe, 
	and link design concepts.  

	The intent is to give the designer the information needed to 
	produce initial designs with better life cycle performance.  

	The tool will be designed to (1) push life cycle issues to the 
	forefront of the design process, (2) provide feedback and related 
	information for each concept to assist in design decisions, (3) 
	structure the concept search and generation so that intent and 
	justifications for concepts are recorded as they evolve, and 
	(4) index the design history using the concept search for retrieval 
	by downstream designers and for archival in the database.  

	Proposed research is focused on determining an appropriate repre-
	sentation for concepts, methods of retrieval from a concept database, 
	and support of design decision making during concept retrieval.  
	
	The target area of application for the prototype is the domain of 
	concepts for packaging of electronics products and mechanical design 
	of computers.

	Alice Agogino is Associate Professor of Mechanical Engineering 
	and Director of the Berkeley Expert Systems Technology Laboratory
	at the University of California, Berkeley, CA. 
67.93William Weihl @CRL, 5/14HERON::ROACHTANSTAAFL !Mon Apr 23 1990 12:2181
Printed by: Pat Roach                                    Document Number: 011468
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Apr-1990 05:12am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: William Weihl @CRL, 5/14

                      Digital Equipment Corporation
                     Cambridge Research Lab Seminar
 
                   Monday, May 14, 1990 3:00pm - 4:00pm
 
		        Multi-Version Memory:
	    Software Cache Management for Concurrent B-Trees
 
			   William E. Weihl
			         MIT
 
Shared-memory and NUMA multiprocessors provide an interface to the programmer
in which there is a single global shared memory (perhaps with some private
memory for each processor).  Such machines typically use fast local memory as
a cache for globally shared data, but the existence of caches is largely
hidden from the programmer.  The abstraction of a single global shared memory
is simple and relatively easy to use, but it hides an important part of the
power of the underlying machine.  In addition, in large-scale multiprocessors,
or in distributed systems, maintaining cache coherence can be expensive.
 
In this talk we argue that it is useful to view cache management as one
specific instance of a more general replication problem.  Furthermore, we
argue that application-specific replication algorithms can give much better
performance than can be achieved by general purpose methods, such as typical
cache coherence algorithms.  Thus, the programmer should be presented with a
model of memory that directly exposes the presence of local cached copies of
data, and allows the programmer to control how these copies are managed.  As
support for our arguments, we describe a new non-standard memory model, which
we call multi-version memory, that could be built given more direct control
over caches.  (It can also be implemented on top of coherent memory, but with
somewhat greater overhead.)  The principal advantage of multi-version memory
is that it allows readers to run in parallel with writers, thus requiring less
delay due to synchronization than coherent memory.  The tradeoff made by multi-
version memory is that it provides a weaker semantics than coherent memory. 
This semantics is still strong enough to be useful: we show how multi-version
memory can be used to implement highly concurrent B-trees.
_______________________________________________________________________________
 
William E. Weihl received the SB degree in Mathematics in 1979, the SB and SM
degrees in Computer Science in 1980, and the PhD degree in Computer Science
in 1984, all from the Massachusetts Institute of Technology.  In 1984 he
joined the faculty of the Massachusetts Institute of Technology, where he is
currently an Associate Professor of Computer Science and Engineering.
 
Dr. Weihl was one of the principal designers of the Argus programming language
and system, which is designed to support the construction of reliable
distributed systems, and the Mercury communications system, which is designed
to support efficient, flexible communication in heterogeneous distributed
systems.  He has made significant contributions to the theory of atomic
transactions, particularly in the design of highly concurrent type-specific
concurrency control and recovery algorithms.  In joint work with Nancy Lynch,
Michael Merritt, and Alan Fekete, he has developed a theoretical framework for
describing and analyzing nested transaction systems.
 
Dr. Weihl's current research interests focus on parallel computing,
particularly in the areas of programming methodology, programming languages,
synchronization, and fault-tolerance.
_______________________________________________________________________________
Digital Equipment Corporation's Cambridge Research Lab is located on the north
side of Technology Square, at the junction of Hampshire Street and Broadway.
The address is:
           Digital Equipment Corporation - Cambridge Research Lab
           One Kendall Square., Building 700, Cmabridge, MA 02139
For questions regarding the seminar, directions, or parking, or to add your 
name to the mailing list, call DTN: 259-6600 or send email to crl:tss.
67.94DR. KUSIAK-4/25-9AMHERON::ROACHTANSTAAFL !Mon Apr 23 1990 12:2351
Printed by: Pat Roach                                    Document Number: 011469
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Apr-1990 05:27am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: DR. KUSIAK-4/25-9AM

	TITLE:	   "Design and Manufacturing:  An AI Perspective"

	SPEAKER:   Dr. Andrew Kusiak
		   Dept. of Industrial & Management Engineering
	           University of Iowa

	DATE:	   Wednesday, April 25, 1990

	TIME:	   9:00 - 11:00 AM

	PLACE:     Pacific Room, DLB12

	HOST:      Tom Cerva
		   AISG Group, AI Technology Center


	In this presentation, an AI approach for solving a number of
	problems in engineering design and manufacturing is discussed.
	The problems considered in this presentation include conceptual
	design of products and systems, group technology, layout design,
	and scheduling of manufacturing systems.

	Dr. Andrew Kusiak is a Professor and Chairman of the Dept. of
	Industrial and Management Engineering at the University of Iowa
	in Iowa City, Iowa.  He holds the Ph.D. degree in Operations
	Research, M.Sc. degree in Mechanical Engineering, and B.Sc.
	degree in Precision Engineering.  Prior to joining the University
	of Iowa he was with the University of Manitoba and the Technical
	University of Nova Scotia.  

	Andrew Kusiak is interested in applications of AI and optimiza-
	tion in modern manufacturing systems.  He is the editor of the
	Journal and Intelligent Systems and the editor of two book
	series, ARTIFICIAL INTELLIGENCE IN INDUSTRY, and APPLIED
	ARTIFICIAL INTELLIGENCE.	
67.95NEXPERT OFFERINGHERON::ROACHTANSTAAFL !Wed May 02 1990 17:3848
Printed by: Pat Roach                                    Document Number: 011532
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     27-Apr-1990 07:04pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEXPERT OFFERING

          PROGRAMMING IN NEXPERT OBJECT TRAINING

          Digital Educational Services offers NEXPERT Object 
          training on a regular basis at the Digital Training Center
          in Marlborough, Massachusetts.

	  Our next offering of this class is May 21, 1990, this
	  is a five day course.  
	  
          ABOUT THE COURSE

          Your job is to implement solutions for real problems.
          Problems involving process control, real-time diagnostics
          and monitoring in manufacturing, design and configuration
          in electronic and mechanical engineering, or classification
          and intelligent decision support in finance and insurance.
          These are real problems you face every day. With
          instruction based upon real-world experience and hands-on
          use of NEXPERT Object, you can gain the skills needed to
          solve real problems with Expert Systems.

          You can ask questions that are most important to your
          business and develop the answers on a Digital VAXstation.

          REGISTRATION, INFORMATION

          For registration call the Educational Services Registration
          office in Bedford, MA:  (617)276-4380.

          For more information on the NEXPERT course, call
          Educational Services at the Digital AI Training Center in
          Marlborough, MA:  (508)490-8893.
67.96REMINDER-DR. MINTON, TOMORROW, 5/30, 9:30AMHERON::ROACHTANSTAAFL !Fri Jun 01 1990 00:3068
Printed by: Pat Roach                                    Document Number: 011668
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     29-May-1990 08:11pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: REMINDER-DR. MINTON, TOMORROW, 5/30, 9:30AM

	TITLE:		"Learning to Improve Problem 
		         Solving Performance"

	SPEAKER:	Dr. Steve Minton
			NASA Ames Research Center

	DATE:		WEDNESDAY, MAY 30, 1990

	TIME:		9:30 - 11:00AM

	PLACE:		Pacific Room, DLB12

	HOST:		Carlos Bhola
			Member of the Technical Staff
			AI Research Group, AITC


 
  	This talk will describe the application of Explanation-Based 
	Learning (EBL) to improve problem solving performance. 

	Previous work has demonstrated that EBL can be an effective 
	method for acquiring domain-specific search control knowledge. 
	However, in practice, EBL may not always generate useful control 
	knowlege.  
	
	For control knowledge to be effective, the cumulative benefits 
	of applying the knowledge must outweigh the cumulative costs of 
	testing whether that knowledge is applicable. Generating effective 
	control knowledge may be difficult, as evidenced by the complex-
	ities often encountered by knowledge engineers. 

	In general, control knowledge cannot be indiscriminately added 
	to a system; it's costs and benefits must be taken into account.
	To produce effective control knowledge, an explanation-based 
	learner must generate "good" explanations -- explanations that 
	can be profitably employed during problem solving.
 
   	The talk will begin with a brief overview of EBL.  Next, the 
	PRODIGY system will be described.  PRODIGY is a planner that 
	learns by searching for good explanations.  Then, the focus 
	will turn to recent work, which includes investigating the use 
	of EBL for complex planning and scheduling applications.
 
	Dr. Steve Minton is currently working in the area of Explana-
	tion-Based Learning with NASA.  He completed his Ph.D. at CMU
	in 1988, under the supervision of Dr. Jaime Carbonell.

	Dr. Minton received the prestigious AAAI Award for best
	scientific paper in 1988, and is viewed as one of the fore-
	most authorities on explanation-based learning within the AI 
	community.  
67.97Some course offerings - July-November, 1990HERON::ROACHTANSTAAFL !Fri Jun 08 1990 11:0780
Printed by: Pat Roach                                    Document Number: 011799
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-Jun-1990 02:34am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Some course offerings - July-November, 1990

From:	SIMVAX::PSMITH "MUSIC, CATS AND LOTS OF GLASS  21-May-1990 1650" 21-MAY-1990 17:31:06.24
Subj:	Mfg. Eng. Training Schedule of Course Offerings

 	MANUFACTURING TECHNOLOGY TRAINING & DEVELOPMENT
              Course/Program Schedule - Q3/Q4 FY'90

This schedule reflects open enrollment offerings from MTT&D located in
the Andover (AET) facility.  Courses for intact work groups are
encouraged and can be arranged at your site at a reduced cost.
Detailed course descriptions and course catalogs are available on
request. Contact SIMVAX::REGISTRAR to register for courses or to obtain
further information.
-------------------------------------------------------------------------------

PROGRAM MANAGEMENT EDUCATION (22 days spread over 6 months)
Begins Oct. 23 - Contact Ray Archambault for further information (SIMVAX::)
_______________________________________________________________________________

K.T. PROBLEM SOLVING & DECISION MAKING  (4 day workshop)  $1000.00
August 21 - 24
October 2 - 5
November 13 - 16
_______________________________________________________________________________

K.T. PROJECT MANAGEMENT   (3 day workshop)   $800.00  
July 17 - 19	September 18 - 20
August 7 - 9	October 30 - November 1
_______________________________________________________________________________

INVESTMENT IN EXCELLENCE     (4 day workshop)     $850.00
July 10 - 13
September 25 - 28
November 13 -16
_______________________________________________________________________________

BECOMING AN EVERYDAY GENIUS   (4 day workshop)   $850.00  
July 31 - August 3	October 16 - 19
August 21 - 24		October 30 - November 2
September 11 - 14	December 4 - 7
_______________________________________________________________________________

SELLING YOUR IDEAS   (3 day workshop)  $750.00                         
July 10 - 12		November 6 - 8
August 14 - 16		December 17 - 19
September 11 - 13
______________________________________________________________________________

MANUFACTURING ENGINEERING FINANCE (3 1/2 day workshop)   $800.00
July 24 - 27                                             $650.00 (Intact Group)	
______________________________________________________________________________

TECHNICAL PROCESS TRAINING:

Design For Manufacturability   (1 day)   $200.00
June 19

Process courses offered by the MTT&D training organization include test
equipment maintenance/programming, soft tools, software simulators,
machine insertion equipment maintenance, surface mount program and 
process. All courses are scheduled as needed and are dependent upon 
availability of the equipment within Manufacturing.  For further 
information contact Paul Sandore (SIMVAX::Sandore).
____________________________________________________________________________
	
67.98NEURON DATA/DEC DEMO DAY 6/8HERON::ROACHTANSTAAFL !Fri Jun 08 1990 11:0854
Printed by: Pat Roach                                    Document Number: 011803
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-Jun-1990 06:35am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEURON DATA/DEC DEMO DAY 6/8

			  NEURON DATA DEC DAY AGENDA

			     FRIDAY, JUNE 1, 1990
		LEOPARD CONFERENCE ROOM/EAGLE CONFERENCE ROOM
			     10:00 A.M. - 2:00 P.M.
			

	      "OVERVIEW OF NEXPERT V2.0 - NEW EXTENDED FEATURES"



    TIME	LOCATION		TOPIC		    SPEAKER
===============================================================================

10:00-10:15	Leopard C.R.	Digital Introduction	    Dino Lachiusa, DEC
		DLB5-3/E10				    3rd Party Relations
							    Mgr.


10:15-10:30	Leopard C.R.	Presentation: Overview	    Nick Halsey, Neuron
				of Nexpert V2.0		    Data
							    DEC Relations Mgr.


10:30-11:00	Leopard C.R.	Presentation: Nexpert	    David Wiser, Neuron
				Execute Library & Special   Data
				Features for VMS & Ultrix   Chief VMS Engineer


11:15-12:00	Eagle C.R.	Demo of Nexpert V2.0	    Nick Halsey
		DLB5-1					    David Wiser


12:00-1:00	DLB5 Cafeteria	Lunch - Open Discussion	    All


1:00-1:45	Eagle C.R.	Optional 2nd Showing of	    Nick Halsey
				Nexpert V2.0 Demo	    David Wiser
67.99FYI: Looks like a good speaker for a seminar - andHERON::ROACHTANSTAAFL !Tue Jun 12 1990 10:5284
Printed by: Pat Roach                                    Document Number: 011891
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Jun-1990 09:49am CET
                                        From:     Pat Roach
                                                  ROACH
                                        Dept:     E/ACT Technology
                                        Tel No:   DTN 828-5784

TO:  Jim Kane                             ( KANE )
TO:  Richard Sitruk                       ( SITRUK )
TO:  Andrew Buchanan                      ( BUCHANAN )


Subject: FYI: Looks like a good speaker for a seminar - and European


    


                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     12-Jun-1990 01:06am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: O-O PROGRAMMING 6/12

        ISB Training and Education                        
     
@@@@@@@@@@@@@@                                                                  
@@ SOFTWARE @@    TECHNICAL SEMINAR                                             
@@@@@@@@@@@@@@                                                                  

TITLE:        How To Teach Object-Oriented Programming To 
              Experienced Software Developers
NUMBER:       SWTS-04                                 
SCHEDULE:     Jun 12, 1990                            
LOCATION:     TBA                                     
TIME:         10:30 a.m.- 11:30 a.m.                           
CLASS SIZE:   50                                      
INSTRUCTOR:   PROFESSOR KARL LIEBERHERR
              NORTHEASTERN UNIVERSITY               
COST:                              $   0                                   
ENROLLMENT CLOSING:                Jun 12, 1990                            
CANCELLATION   IT IS NOT NECESSARY TO HAVE MANAGER APPROVAL TO ATTEND.     
POLICY:        REGISTRATION IS "ONLY" USED FOR PROJECTED ATTENDANCE.       
ENROLLMENT RETURN TO:    Registrar                               
                         HYEND::REGISTRAR          
                         DTN: 297-5097 OR                           
                         ON LINE:  $COURSES
                         PASSWORD: $TRAINING
                         SELECTION: SOFTWARE TRAINING                      
                                                                                
ABSTRACT:  During the last 5 years we have developed an object-                 
oriented design method (the Demeter method) which is programming                
language independent and we have developed a tool (the C++ Demeter              
system) which supports the method for C++. The method is based                  
an class dictionaries which describe the high-level relationship                
between classes.                                                                
                                                                                
We will describe the Demeter method and how we use it to teach                  
object-oriented programming to experienced software developers.                 
                                                                                
PRESENTOR: Karl J. Lieberherr got his Ph.D. at the Swiss Federal                
Institute of Technology in Zurich, Switzerland and was Assistant                
Professor of Computer Science at Princeton University.  Before                  
joining Northeastern University in 1985 as a Professor, he was                  
Principal Member of Technical Staff at GTE Laboratories.                        
                                                                                
During the last 5 years his research team in object-oriented                    
software engineering has developed the Demeter method for                       
object-oriented design and programming and his team has built                   
a supporting CASE tool, called the C++ Demeter system. Some key                 
contributions of the Demeter project are: the concept of a class                
dictionary, the Law of Demeter, the application planning technique              
and the class restructuring algorithms.                                         
67.100HUMAN INTERFACE-6/25-MLOHERON::ROACHTANSTAAFL !Wed Jun 13 1990 17:36157
Printed by: Pat Roach                                    Document Number: 011924
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     13-Jun-1990 02:37pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: HUMAN INTERFACE-6/25-MLO


			   HUMAN INTERFACE SEMINAR

			    Monday, June 25, 1990
		      General Doriot Auditorium, MLO4-5

Sponsored by: Corporate Research & Architecture, Technology Transfer Program



				   AGENDA
 9:00	Welcome                                           Ken Beers/Bill Kuhlman

 9:10   Overview of Human Interface Research at MCC	  Bill Curtis

 9:20   Human Interface Tool Suite (HITS) and Runtime     David Wroblewski
        Architecture

 9:50   Work Coordination Through Action Recommendations  David Wroblewski
        from Interface Objects

10:10   An Object-Oriented Language for Interactive       Mark Tarlton
        3-D Visualization

10:30   Break

10:40   Motivation for Stylus Based Interfaces and        Gale Martin
        Handwriting Recognition Technologies

11:30   Spatial Parsing for Visual Interface Languages    Kent Wittenburg
 
12:00   Wrap-up and Technology Transfer Issues

12:30...Lunch with Speakers and Follow-up Interactions
        (Russo Conf. Rm., MLO1-3, pole 1C)  



				ABSTRACTS

 
HUMAN INTERFACE TOOL SUITE (HITS) AND RUNTIME ARCHITECTURE  and...
WORK COORDINATION THROUGH ACTION RECOMMENDATIONS FROM INTERFACE OBJECTS
(David Wroblewski)
In these talks, I will discuss two components of the ongoing HITS work.  First, 
the HITS blackboard, which provides the framework in which multiple knowledge 
sources can be used to integrate and react to multimodal input.  The second 
component is called DETENTE, and has the goal of providing practical support 
for representing and managing representations of user activities.  I will 
examine the motivation behind our technical approach to DETENTE, and examine 
a number of interfaces that exhibit DETENTE-like qualities.

AN OBJECT-ORIENTED LANGUAGE FOR INTERACTIVE 3-D VISUALIZATION
(Mark Tarlton)
Traditionally, computer graphics has been approached as a programming problem 
where interactive graphical presentations are produced by writing sequential, 
procedural programs.  These procedures invoke drawing functions from a graphics 
library to form the images on a display.  While there are a number of problems 
that arise in taking this approach, the most obvious is that substantial 
expertise in both graphics and programming are required to produce even static 
3-D images. When interactive and dynamic features are added to the presentation,
the task becomes daunting, even for experienced developers. Our hypothesis is 
that many of the problems that face the developers of interactive 3-D systems 
result from using a procedural specification of graphic operations.  The 
language I will present allows the specification of interactive, animated 
graphical presentations through the construction of object graphs.
	
MOTIVATION FOR STYLUS BASED INTERFACES AND HANDWRITING RECOGNITION TECHNOLOGIES
(Gale Martin)
I will describe the vision fueling MCC's Interactive Worksurface Project and the
technologies we are developing to make this vision a reality.  This includes 
general overviews of the hardware for interactive tablets (large and small),
software drivers, handwriting recognition capabilities, and techniques for
interpreting the visual languages communicated on interactive tablets.

SPATIAL PARSING FOR VISUAL INTERFACE LANGUAGES
(Kent Wittenburg)	
The overall goal is to apply a language perspective to the problem of 
connecting meaning to graphical and sketched media on both the input and the
output side of human-machine interfaces.  The technology will initially be 
applied to the problem of recognition of hand-sketched input through the 
electronic paper interface of the interactive worksurface (IWS).  Our first 
application will be a math expression interpreter for an electronic math sketch
pad.  Subsequent applications will include interpretation of sketched designs
(e.g., engineering or architectural layouts) in such a way that the semantic 
information can be made available for subsequent querying, intelligent 
advising, creation of dynamic prototypes, etc.  On the output side, we expect 
that the inverse connection of underlying data to a visual vocabulary will 
enable easy-to-use tools for connecting the semantics of underlying data to 
dynamic graphical displays.




			  ABOUT THE SPEAKERS

BILL CURTIS is currently the Director of MCC's Human Interface Laboratory where 
he directs research on developing a new generation of user interface 
technology.  Previously Dr. Curtis was a Director in MCC's Software Technology 
Program, which is developing a new generation of technology for supporting the 
design of large systems.  Prior to joining MCC, Dr. Curtis was the Manager of 
Programming Trends Analysis at ITT's Programming Technology Center; Manager of 
Software Management Research in the Space Division of General Electric, and 
Research Assistant Professor at the University of Washington.  He has published 
over 80 articles on software engineering, human-computer interaction, and the 
management of large systems development, and has edited two books: Human 
Factors in Software Development an IEEE tutorial (2nd ed., 1985) and Human 
Factors in Computing Systems - II published by Elsevier (1985). He is on the 
editorial boards of the IEEE Transactions on Software Engineering, 
Human-Computer Interaction, the International Journal of Man-Machine Studies, 
and the Journal of Systems and Software.

DAVID WROBLEWSKI is a Member of the Technical Staff at MCC's Human Interface
Laboratory, where he has worked since 1985, participating in projects ranging 
from natural language analysis to planning for intelligent advising to his 
current research on supporting human work through embedded agendas.  Prior to 
MCC, he worked for 4 years in the Honeywell Systems and research center doing 
expert systems and autonomous vehicle research and in production software 
engineering.  Mr. Wroblewski holds a B.S. in Computer Science from the 
University of Wisconsin.

MARK TARLTON is a a Ph.D. candidate in Computer Sciences at the University of
Texas.  For the last five years in the MCC Human Interface Lab, he has been 
involved in the development of "Pogo", an object-oriented platform for dynamic, 
two and three dimensional graphical interfaces.  Previously, he was a research 
assistant at the University of Kansas Medical Center where he conducted 
research in medical applications of interactive computer graphics and image 
processing.  He is the author of more than 25 technical papers in the area of 
visualization and interactive graphics.

GALE MARTIN received her Ph.D. in experimental psychology from the University
of California in 1981.  Her dissertation research dealt with human visual
pattern recognition.  Since then she has worked for Eastman Kodak, and for the
past five years has been assigned as a Kodak representative at MCC.  She
has been active in neural network research focusing on handwriting recognition.

KENT WITTENBURG is the leader of the Interface Languages research in the MCC 
Human Interface Lab.  He holds a Ph.D. from the University of Texas at Austin, 
where he studied natural language parsing and Categorial Unification Grammar.  
He has previously held computational linguistics positions at SRI 
International and the Linguistics Research Center at the University of Texas.
67.101FWD: DCVN BROADCAST INFORMATIONHERON::ROACHTANSTAAFL !Wed Jun 20 1990 15:19384
Printed by: Pat Roach                                    Document Number: 012012
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Jun-1990 01:55pm CET
                                        From:     PAPAGEORGE
                                                  PAPAGEORGE@AITG@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: FWD: DCVN BROADCAST INFORMATION

From:	NUTMEG::NUTMEG::MRGATE::"A1::MOLLOY.MARY"   19-JUN-1990 17:06:08.90
To:	@Distribution_List
CC:	
Subj:	DCVN BROADCAST INFORMATION

From:	NAME: MARY MOLLOY  @TTB             
	FUNC: MDP                             
	TEL: 264-3942                         <MOLLOY.MARY AT A1 at NUTMEG at 
TTB>
To:	See Below


    Please see the attached memo.
    
    Thanks.
    
    Regards,
    Mary
    




From:	NAME: SOPHIA JOHNSON                
	FUNC: DIGITAL VIDEO NETWORK           
	TEL: 276-4089/DTN249-4089             <JOHNSON.SOPHIA AT A1 at systmx at 
BUO>
Date:	19-Jun-1990
Posted-date: 19-Jun-1990
Precedence: 1
Subject: DECWORLD BROADCAST REPORT
To:	See Below
CC:	See Below




To Distribution List:

PIZ@EXIT26@VMSMAIL,
NAME: DON ELIAS <ELIAS.DON AT A1 at systmx at BUO>,
NIKKI RICHARDSON @MLO,
_ASABET::SHIPMAN AT A1 at systmx at BUO,
_EXIT26::RUDOMIN AT A1 at systmx at BUO,
_EXIT26::DAHL AT A1 at systmx at BUO,
_EXIT26::MASLOWSKII AT A1 at systmx at BUO,
MARY MOLLOY @TTB,
GARY KUBA @OGO

CC Distribution List:

MEADE@EXIT26@VMSMAIL




Author:	SOPHIA JOHNSON                
Date:	23-May-1990
Posted-date: 19-Jun-1990
Precedence: 1
Subject:                                                                         




6/19/90			DCVN  AT  DECWORLD  BROADCASTS
 

1. JULY 10

   The Vision of Digital's Style of Computing  		(WILLIAM STRECKER)
							(@MKO, 223-3726)

	   John Kunis @OGO, 276-0988/Peter Damon @OGO, 276-8786

*BOB BICKFORD		BUZZ LUTTRELL-MODERATOR			MEMO SENT  6/6
================================================================================
2. JULY 11

(S-425-0-001)   Achieving Your Business Goals Through Software Integration
			DRI: Dennis Phelan, @ttb, 264-2030
			Speaker:  Jack Walsh (OUTSIDE CONSULTANT)
					
(S-425-0-002)   NAS:  The Perfect Solution for Your Multi Vendor World
			DRI:  Dennis Phelan
			Speaker:  Jackie Kahle, ttb, 264-3988
			Backup:  Joanne MacMullen
							(SHARON GUADAGNO @TTB
							 (264-1814) 
							(BARB LIBERTY @TTB
							 (264-2591)
								MEMO	5/29
*ERIC BASTA		ANN MCGRATH - MODERATOR
=============================================================================== 

3. JULY 12

(S-415-0-001)   Digital's Office Strategy: Enterprise Resources for the 
                End-User 
			DRI:  Ken Walker, TTB 264-3711
			Speaker:  Gene Hodges,TTB 264-3734
			Backup:  Chris Reed, Steve Kirchoff,TTB 264-4384
           		Murray Copp, TTB 264-3328, Joanne Masingill, TTB
								264-3122

(A-415-0-004)   ALL-IN-1 Phase II:  Integrated Desktop Applications for
                Today and Tomorrow
			DRI: Ted Kennedy, TTB 264-3389
			Speaker: Steve Kirchoff
			Backup: Steve Martin, Joanne Correia
				TTB 264-2575 / TTB 264-0943
							(KEN WALKER)
PAM PORTER-ELLIOTT	MARSHA PITTS-WILLIAMS - MOD	(@TTB, 264-3711)
(603)878-1049							MEMO 6/18
================================================================================




4. JULY 13

(A-405-0-001)   Information Systems (IS) Planning for the Enterprise (2nd)
			DRI: Jack Richardson
			Speaker: Dave Caperelli
			Backup: Roger Greeley, Nancy Settle, Bart Bolton
		
(A-405-0-009)   From Chaos to Architecture 
			DRI: Jack Richardson
			Speaker: Jim Friel
			Backup: Henry Therberge, Bud Gorman
 			        Worth Robbins, Ken Gordon		
(alternative)
   
(A-405-0-002)   Digital's Answer to Computing for a Changing World (4th)
			DRI: Jack Richardson
			Speaker: Clare Peiffer
			Backup: Bart Bolton, Mike Schiff, Chad Cutler
			        Henry Morris
							(BOB WEINER @PDM
							 (291-0474)
*Jennifer Abod		JACKY HOVEY - MODERATOR          memos sent 5/31
================================================================================
5. JULY 16	

   (A-420-0-001)   Moving Ahead of Your Competition with Business and
                Technical Documentation Solutions
			DRI: Shelley Spierer, @TTB, 264-0008
			Speaker: John Giudici, @TTB, 264-2238
		
  (D-420-0-001)   Document Management - Staying Afloat in the Flood
                  of Electronic Documents
			DRI: Susan Kaufmann, TTB, 264-5308
			Speaker: Geoff James, @TTB, 264-1490 
		
  (D-420-0-003)   Eliminating Computer Babel with the CDA Architecture (TM) 
			DRI: Tom Azcona, @TTB, 264-3451
			Presentor: Tom Azcona/John Clancy, @TTB, 264-3982
				   John Hill
							(SHARON GUADAGNO @TTB
	               CC: JOANNE LUTIAN @TTB		 264-1814)
							(HOWARD WOOLF @TTB)
							 264-3901)
*ERIC BASTA		(STEVE SINE - MODERATOR)		MEMO    5/29
================================================================================



6. JULY 17

   Secure the Future:
   Digital's Approach to Information Security		(STEVE Redfern)

*BOB BICKFORD @CTS	(ANN MCGRATH - MODERATOR)		  MEMO   5/29	

================================================================================
7. JULY 18 

(S-430-0-001)   Strategic Software Development for Success      
			DRI/presenter:  Marion Dancy @PDM, 291-0138
			Backup: Geoff Roach, @PDM, 291-0462
		
(A-430-0-001)   IS Application Development: From Chaos to Control
			DRI & Speaker: Shirley Ann Stern, @PDM, 291-0426
			Backup: R.J. Mathew, @PDM, 291-0313
							      (SUE APPEL @PDM
*GLENN LITTON		(MARCIA PITTS-WILLIAMS - MOD)		291-0458) or
(617) 661-4828						      (508)490-0458
================================================================================
8. JULY 19

(S-410-0-001)   Digital and UNIX (r): Strategy & Products
			DRI:  Roy Shiderly
			Speaker: Joe Menard 
			Backup:  Rick Berzle
		
(A-410-0-001)   Success with Open Systems
			DRI:  Roy Shiderly
			Speaker: K. Comisky/Jim Isaak
		
*FRANCINE CAMPBELL					(JOE MENARD @ZKO
							 381-2929)
			(BUZZ LUTTRELL - MODERATOR)		MEMO     5/29
===============================================================================	
9. JULY 20

(A-405-0-007)   Digital's "Information Network": Dealing with the Data Dilemma
			DRI: Chuck Rozwat, @NUO, DTN 264-6104
			Speaker:
			Backup: Wendy Caswell, @NUO, DTN 264-2833
		
(A-405-0-003)   Making the Right Database Choice
			DRI: Vickie Farrell, @NUO, DTN 264-2412	
					   		     (BOB WIENER @PDM
								 291-0474)
								UPDATED 6/14
*MIKE GOFF      	(PEG O'BRIEN - MODERATOR)		MEMO   6/1
================================================================================


10. JULY 23

(A-405-0-008)   Gaining A Competitive Advantage With DECtp  			
			Dri: Peter Powell
			Speaker: Peter Powell, John Anderson, Roger Baust
                                 (all primary speakers, will rotate)
			Backup: Peter Powell, John Anderson, Roger Baust
		
   Integrated Production Systems: Solutions for Mission-Critical Applications	
	Jack Lucier @MR0, 297-4261 / Cathy Stockwell @MRO, 297-5347
							(BOB WIENER @PDM
							 291-0474)
*ERIC BASTA		(JACKIE HOVEY - MODERATOR)	           MEMO  5/29
================================================================================
11. JULY 24

    Investment Criteria for the 90's: A Comprehensive Approach to 
    Cost of Ownership

	      Rene Martinez/Lorainne Tartaglia/Index Group - Consultant
	     PKO, 223-3405 /CYCLPS::Tartaglia

    VAXComputing: Protecting Your Investment		
			Jim Brierley/Rene Martinez
		       BXB, 293-5265/PKO, 223-3405  

*LISA RYAN	       (ANN MCGRATH - MODERATOR)		 MEMO  5/29	
===============================================================================
12. JULY 25
 
	Digital : the Performance Leader in Open Networks
			 BOB (BJ) Johnson/Gail Daniels
			LKG, 223-7725    /LKG, 226-5169

	Using Open Networks in the Information Systems Environment
							(DENNIS ROSSITER @MK0-
							 BILL MITCHELL @LKG)
			Mike Gayowski/Bill Mitchell      JEFF LUKOWSKY @LKG
			MKO, 264-3716/LKG, 226-7884

*FRANCINE CAMPBELL	(BUZZ LUTTRELL - MODERATOR)		MEMO  5/29
================================================================================
13. JULY 26
	 
    Systems Integration: Honing Your Competitive Edge
							(LEE GALLANT, MRO
							 297-3858.
							CC: JOHN O'DONNELL @MRO

    Strategic Investments: Qualifying their Impact on your Bottom Line		
							 LYNDON WILKES @MET
*SUSAN SIROIS					        
			(MARSHA PITTS-WILLIAMS - MOD)		MEMO SENT 6/6
================================================================================


14. JULY 27
 
     Desktop Integration in a Multi-vendor Environment
 			  Jim Deasy/Bruce Hunter
		     LJO, 226-2450/MLO, 223-1869

    Client/Server computing: Products and Approaches for Today
			  Steve Koenig/Joan Ross	(JIM DEASY)
			BXB, 293-5670 /PKO, 223-6213

*FRANCINE CAMPBELL	(JACKY HOVEY - MOD)			MEMO  5/29
================================================================================
15. JULY 30

    The Ultimate Data Center: Achieving Maximum Productivity
			  Jack Lucier/Brian Maher
			MRO, 297-4261/MRO, 2977085

    VAX Super Computing: Solutions to Complex Problems
			 Pamela Eastman/Dave White
			BXB, 293-5670  /MRO, 297-5844

*LISA RYAN		(ANN MCGRATH - MOD)			MEMO   6/5
================================================================================

16. JULY 31

    Can you Afford Business as Usual: An Introduction to DECimage

    Making Image Capability Available across the Enterprise

    Content Driver - Barry Washington,  DTN 264-1077, VISUAL::Washington

	Presentors:  Joel Arker		264-1696 @ MK0
		     Bruce MacDonald    264-1076	
		     Ron Higgins	264-0868
		     Richard Kalin	264-6145
								UPDATED 6/15
*MIKE GOFF	     (BUZZ LUTTRELL - MOD)			MEMO SENT 6/7

================================================================================
* PRODUCERS

To Distribution List:

NAME: Howard Woolf <WOOLF.HOWARD AT A1 at NUTMEG at TTB>,
NAME: Gene Hodges <HODGES.GENE AT A1 at NUTMEG at TTB>,
NAME: Michael Carabetta @TTB <CARABETTA AT A1 AT SAGE AT TTB>,
NAME: Stephen J. Hort @TTB <HORT.STEPHEN AT A1 at NUTMEG at TTB>,
NAME: Bob Crawford @TTB <CRAWFORD.BOB AT A1 at NUTMEG at TTB>,
NAME: Dick Rose @TTB <ROSE_DICK AT A1 at NUTMEG at TTB>,
WEINER @CSGDEC @MRGATE,
STEVE THOMAS @MRO,
NAME: Tom Richardson <RICHARDSON.TOM AT A1 at NUTMEG at TTB>,
CLAIRE MESSIER @TTB,
MEL RAY @MRO,
CATHY LANGE @MEL,
DANCY @CSGDEC @MRGATE,
HESS @CSGDEC @MRGATE,
NOVOSON @CURIE @VAXMAIL,
PORELL @CSGDEC @VAXMAIL,
EPSTEIN @CSGDEC @VAXMAIL,
GARY W HOPPE @MRO,
HORNBACH @SSGBPM @VAXMAIL,
SLATTERY @SSGBPM @VAXMAIL,
RICHARDSON @CISM @VAXMAIL,
MULLEN @CISM @VAXMAIL,
PAPAGEORGE @AITG @VAXMAIL,
AUBERG @SSGBPM @VAXMAIL,
MENARD @XIRTLU @VAXMAIL,
LMOORE @CURIE @MRGATE,
GRENHAM @TPSYS @VAXMAIL,
NAME: HENRY ANCONA @TTB <ANCONA AT A1 AT SAGE AT TTB>,
JANE LONGPRE @AKO,
KEN WALKER @TTB,
NAME: Barbara Liberty <LIBERTY.BARB AT A1 at NUTMEG at TTB>,
DICK MOLLIN @MLO,
DOHERTY @CURIE @MRGATE,
NAME: Gia Milo @TTB <MILO.GIA AT A1 at NUTMEG at TTB>,
GILLIGAN @BUFFER @VAXMAIL,
NAME: Henry Ogilby @TTB <OGILBY_HENRY AT A1 at NUTMEG at TTB>,
ABERNETHY @SKETCH @VAXMAIL,
PAM JOHNSON @MLO,
JIM WHITE @ALF,
SHIDERLY @XIRTLU @MRGATE,
NAME: Michael Smith <SMITH.MIKE AT A1 at NUTMEG at TTB>,
NAME: Jeanne Marquis <MARQUIS_JEANNE AT A1 at NUTMEG at TTB>,
NAME: SYDNEY STEVENS @TTB <STEVENS.SYDNEY AT A1 at NUTMEG at TTB>,
ARREL WHITE @MEL,
dottie tasker @mel,
NAME: Joan Johnson @TTB <JOHNSON_JOAN AT A1 at NUTMEG at TTB>,
HAMMOND @CSGDEC @VAXMAIL,
NAME: ANDY COPANAS  @TTB <COPANAS.ANDY AT A1 at NUTMEG at TTB>
67.102Dr.Salvendy-6/29-9:30HERON::ROACHTANSTAAFL !Fri Jun 29 1990 18:5771
Printed by: Pat Roach                                    Document Number: 012141
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     28-Jun-1990 06:33pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Dr.Salvendy-6/29-9:30





	TITLE:		"The Future of Human Aspects in Computing"

	SPEAKER:	DR. GAVRIEL SALVENDY
			Purdue University

	DATE:		Friday, June 29, 1990

	TIME:		9:30 - 11:00 AM

	PLACE:		Pacific Room, DLB12

	HOST:		MITCH TSENG
			Intelligent Systems Technologies Group
			AI Technology Center


	An effective, productive and humane framework for designing,
	implementing and using advanced computerized systems will be
	presented.  

	Within this framework, research studies are outlined in the 
	areas of knowledge structure, learning knowledge elicitation, 
	interface design and personnel scheduling, which results in 
	increased effectiveness of knowledge-based systems and inter-
	face functioning.  

	The seminar will conclude with the implications of these 
	studies for the design of adaptive interfaces, modulized
	flexible automated production of knowledge-based systems
	and the design of organizational learning.

	Gavriel Salvendy is a member of the National Academy of
	Engineering.  He is the author or co-author of over 200
	publications, including ten books.  Many of his publications
	have been translated into Japanese, Chinese, Russian, Spanish,
	Serbo-Crotian and Portuguese.  

	Dr. Salvendy is the founding editor of the international 
	Journal on Human-Computer Interaction and the International 
	Journal on Human Factors in Manufacturing; Founder and Chairman 
	of the International Commission on Human Aspects in Computing, 
	headquartered in Geneva, Switzerland.  

	In 1984 he was appointed to a personal chair established by 
	the NEC Corporation of Japan specifically for Dr. Salvendy.  
	He is twice recipient of the Fulbright "Distinguished Professor 
	Award".  His main research interests are in the human aspects  
	of design and use of advanced technologies.

	NOTE:	This seminar WILL BE videotaped.
	
67.103NEXPERT Training, 7/16HERON::ROACHTANSTAAFL !Fri Jun 29 1990 19:0845
Printed by: Pat Roach                                    Document Number: 012070
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Jun-1990 03:33am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: NEXPERT Training, 7/16

          PROGRAMMING IN NEXPERT OBJECT TRAINING

          Digital Educational Services offers NEXPERT Object 
          training on a regular basis at the Digital Training Center
          in Marlborough, Massachusetts.

	  Our next offering of this class is July 16, 1990, this
	  is a five day course.  

	  ABOUT THE COURSE

          Your job is to implement solutions for real problems.
          Problems involving process control, real-time diagnostics
          and monitoring in manufacturing, design and configuration
          in electronic and mechanical engineering, or classification
          and intelligent decision support in finance and insurance.
          These are real problems you face every day. With
          instruction based upon real-world experience and hands-on
          use of NEXPERT Object, you can gain the skills needed to
          solve real problems with Expert Systems.

          You can ask questions that are most important to your
          business and develop the answers on a Digital VAXstation.

          REGISTRATION, INFORMATION

       	  For more information on the NEXPERT course,or registration call
          Educational Services at the Digital AI Training Center in
          Marlborough, MA:  (508)490-8893.
67.104EPITOOL OFFERING 7/23-27HERON::ROACHTANSTAAFL !Fri Jun 29 1990 19:0878
Printed by: Pat Roach                                    Document Number: 012068
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Jun-1990 00:32am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: EPITOOL OFFERING 7/23-27

             ******************************************************
             *                                                    *
             *       Educational Services Intelligent Systems     *
             *                                                    *
	     *		      is happy to announce                *
             *                                                    *
             *            the next offering of EPITOOL            *
             ******************************************************

Dates:		July 23-27, 1990

Location:	Caribbean Classroom - DLB12-1

Registration:	To register for this course call Christine Johnson at
		291-8893.

Information:	For more information contact 
		Malin Goodwin at ESIS::GOODWIN or DTN 291-8894	

About EPITOOL:
--------------
	EPITOOL is a Lisp based hybrid expert system tool developed by 
	Epitec AB in Link�ping, Sweden. As announced at IJCAI in Detroit,
	EPITOOL will be a software product distributed and sold by Digital.

	EPITOOL	has a wide range of knowledge representations including 
	forward and backward chaining rules, objects and methods, and 
	functions for represenation of procedural knowledge. Epitool also
	provides functions for creation of Decwindows-style application
	graphics, and call-in  call-out to any VAX supported language.

	Development of Knowledge bases can either be done on a VAX station,
	using development graphics, windows and menus, or on a regular
	terminal using Epitools own command language.  
	
About the course:
-----------------

	The course consists of lectures and labs. Labs will provide the 
	students with a hands on experience working with the tool. The
	topics covered in the course are:

	- Knowledge Representation in Epitool: 
		Concepts, Aspects, Individuals, Objects-Methods.

	- Actions:
	  	Rules, Backward and Forward chaining. Functions.
		Rule sets.

	- Development tools:
		Error handling, Debugging

	- Values:
		Uncertain, Unknown, Descriptions, Groups.

	- End user interface:
		Questions, Application graphics, Explanations,
		Streams and active paths

	- Integration

	- Customizing the development environment.
67.105VAX Decision Expert 7/23-27HERON::ROACHTANSTAAFL !Fri Jun 29 1990 19:0941
Printed by: Pat Roach                                    Document Number: 012067
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     23-Jun-1990 00:02am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: VAX Decision Expert 7/23-27

PROGRAMMING IN VAX DECISION EXPERT

Digital Educational Services offers Programming in VAX 
Decision Expert training on a regular basis at the Digital AI 
Training Center.  Our next offering of this class is July 23-27,1990
in the Arctic Classroom.

ABOUT THE COURSE

This course is designed to train non-AI programmers and 
engineers in the use of VAX Decision Expert in building Expert 
Systems.  The course will focus on the use of VAX Decision 
Expert to build applications in diagnostics, maintenance and 
decision trees, although it can be used in a variety of other areas.  
Working on VAXstations students will have the opportunity to 
develop practical skills using both VAX Decision Expert's 
development and delivery environments.

REGISTRATION INFORMATION

For registration call Christine Johnson at 291-8893.

For more information on the course, call Educational Services at 
the Digital AI Training Center, Marlborough, MA:  
(508)490-8893.
67.106PROF. KOLODNER-7/25-10AMHERON::ROACHTANSTAAFL !Wed Aug 01 1990 12:29115
Printed by: Pat Roach                                    Document Number: 012309
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     17-Jul-1990 05:44am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI@VBO
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PROF. KOLODNER-7/25-10AM





	TITLE:		"A Case-Based Approach to Solving
			Complex Real World Problems"

	SPEAKER:	PROFESSOR JANET L. KOLODNER
			Georgia Institute of Technology 

	DATE:		Wednesday, 25 July, 1990

	TIME:		10:00 AM - NOON

	PLACE:		Pacific Room, DLB12

	HOST:		JEFF CLANON
			AI Technology Center Training Manager


	In case-based reasoning, new problems are solved by remembering
	(retrieving) previous problem situations similar to a new one,
	and adapting retrieved solutions to fit the new problem. Case-
	based reasoning is useful for design tasks, planning, diagnosis
	problems, and common-sense problem solving.  It is an inference
	method people use quite often in their day-to-day reasoning for
	both expert and common-sense tasks, and it provides an alternate
	way of building expert systems.

	Research in automating case-based reasoning has focused mostly
	on retrieval and adaptation issues.  Case-based systems have
	been developed that resolve arguments, create recipes, plan
	meals, design landscapes, configure machines, etc.  Several
	systems are in use in industrial and military settings, and 
	are performing quite admirably.

	In this seminar, case-based reasoning will first be described, 
	spending the most time presenting case retrieval.  Then, 
	there will be a discussion as to what makes these systems work
	well and what is needed to extend them to work on more complex
	problems.  

	The conclusion presented is that they have to be made to work
	along with people to solve problems, with the computer augument-
	ing human memory and the person doing novel adaptations and 
	making esthetic and other judgments.  Such systems are called
	{\em case-based decision aiding systems}.

	Several real and idealized systems will be presented to illus-
	trate the responsibilities of machines and people working
	together.  Then, guidelines will be presented for building
	such systems.  In conclusion, their potential benefits will be
	summarized.





	Professor Janet L. Kolodner
	25 July, 1990 - Page Two



	Janet L. Kolodner is a Prof. in the College of Computing at
	Georgia Institute of Technology.  She received her Ph.D. in
	computer science from Yale University in 1980.

	Prof. Kolodner's research work investigates issues in learning,
	memory and problem solving.  As part of these investigations,
	she has pioneered a reasoning method called case-based reasoning.
	
	In case-based reasoning, the results of previous cases are 
	applied to new cases, cutting down the complexity of the 
	reasoning necessary in the later cases and allowing a problem
	solver to anticipate and avoid previously-made mistakes.  

	Case-based reasoning projects from prof. Kolodner's lab include
	MEDIATOR and PERSUADER, common sense and expert mediation pro-
	grams; JULIA, a case-based design problem solver; MECH, a case-
	based logistics manager.  In recent work, she is investigating 
	the implication of the cognitive model implied by the case-based 
	reasoning paradigm in the areas of decision aiding and education.

	Prof. Kolodner has written one book and edited two.  RETRIEVAL AND
	ORGANIZATIONAL STRATEGIES IN CONCEPTUAL MEMORY:  A COMPUTER MODEL
	describes a memory model that integrates semantic and episodic
	memory and provides cognitively-plausible strategies for searching
	the memory.  It also presents CYRUS, a computer program that 
	implements the model and acts as an intelligent fact retrieval
	system.  

	MEMORY, EXPERIENCE, AND REASONING, an edited volume, explores 
	the issues associated with giving reasoning systems the capability
	of making use of their past experiences.  

	PROCEEDINGS:  CASE-BASED REASONING WORKSHOP is a collection of
	papers describing the state of Case-Based Reasoning in 1988.  She
	is currently working on a case-based reasoning textbook, and has
	authored some dozens of technical papers.
67.107K-B SYSTEMS OPENINGSHERON::ROACHTANSTAAFL !Fri Aug 17 1990 17:0568
Printed by: Pat Roach                                    Document Number: 012631
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     17-Aug-1990 00:12am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: K-B SYSTEMS OPENINGS

TO REGISTER:    Please send mail to Suegene Levin at AIADM::LEVIN.  Be sure to
	      	specify which date you prefer.


********************************************************************************
		OVERVIEW OF AI/KNOWLEDGE-BASED SYSTEMS
********************************************************************************


DESCRIPTION:	This is a one-day overview of the fields of Artificial 
		Intelligence (AI) and Knowledge-based Systems (KBS).  It is
		presented by MIT Professor Randall Davis, a world renowned
		leader in the field. Professor Davis takes the "mystery" out of
		AI and explains how Knowledge-based Systems can solve business
		problems differently than conventional systems. In addition,
		he discusses the impacts of technical, business and
		organizational issues on implementing knowledge-based systems.

		Also included in this seminar is a short presentation of
		Digital's involvement in AI.

GOALS:		Participants will be able to:

		o Identify the different technologies that comprise the 
		  field of AI.

		o Identify applications that use AI technology.

		o Discuss some of the issues involved in selecting an 
		  appropriate application for Knowledge-based Systems 
		  technology

		o Understand some of the future directions of AI technology.

PRESENTERS:	Professor Randall Davis, MIT

OFFERINGS:	September 10, 1990
	        October 29, 1990
	        January 7, 1991
		February 25, 1991
		April 29, 1991

LENGTH:		1 day

FORMAT: 	Lecture
 
COST:		$500.

LOCATION:	Pacific Conference Room, DLB12 
		295 Donald Lynch Boulevard
		Marlborough, MA  
	
67.108AICorp KBMS TrainingHERON::ROACHTANSTAAFL !Mon Aug 27 1990 17:4142
Printed by: Pat Roach                                    Document Number: 012823
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Aug-1990 10:44pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: AICorp KBMS Training

PROGRAMMING IN KBMS 

AICorp's 3-day training on KBMS Application Development.  This
course introduces Digital to AICorp's Knowledge Based Management 
System (KBMS) tool recently ported-to the VAX/VMS environment.
This is the first offering to Digital Employees.  The course 
begins Tuesday Sept 4, 1990 at the Digital AI Training Center
in Marlborough Mass.

ABOUT THE COURSE

This course is for programmers or knowledge engineers who want to 
learn to develop knowledge base applications using KBMS.  Topics
include knowledge base systems and conventional programming 
techniques, KBMS components and inferencing strategies, KBMS 
rule syntax, features for enhancing fluency and rule definition,
and the KBMS development cycle.  The class uses KBMS for 
DECwindows to demonstrate product features.  Students will have 
the opportunity to develop several small applications using
both graphic and non-graphic techniques.

REGISTRATION

For registration or more information call Christine Johnson, 
Educational Services at the Digital AI Training Center, 
Marlborough, MA, DTN 291-8893.
67.109Message from AI TrainingHERON::ROACHTANSTAAFL !Mon Aug 27 1990 17:43350
Printed by: Pat Roach                                    Document Number: 012824
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     22-Aug-1990 11:29pm CET
                                        From:     MALLORY
                                                  MALLORY@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Message from AI Training

	To the AITC:

	The AI Training and Education group would like to inform the AITC
	of the changes that have occurred within our programs.  Also, this 
	is an excellent opportunity to send out our FY91 schedules!
	
	The Knowledge Engineering Curriculum is now a standalone product.
	We hope that we are able to hit a training market area and perhaps
	generate business for SELECT and/or Fellowship.  The AITE is also 
	gearing up to deliver a Methodology Curriculum (a limited number) at
	customers site within six months.  

	The Intro to AI in Business module has a title change.  "Introducing 
	and Transferring AI Technology" better reflects the course intent;
	incorporating Technology Transfer as a major theme of this four-day 
	course.  Day 1, or "Overview of AI and Knowledge-Based Systems," has 
	been re-structured by Randy Davis of MIT.  Prof. Davis will speak on 
	how knowledge-based systems can solve business problems and the impact 
	of technical, business and organizational issues have on implementing 
	knowledge-based systems.

	The Advanced Curriculum has also taken on a new name - "Emerging
	Software Technology Training."  Two new courses offered and presented 
	by CGI this Fall are a two-day KnowledgeCraft Hands-On Introductory 
	Seminar and a two-day PAKS Seminar.  AITE has developed a new course 
	for employees and our customers entitled Designing, Integrating and 
	Testing KBS - details to follow at a later time.
	
	The attached pages reflect the three programs AITE offer - Knowledge
	Engineering, Emerging Software Training Technology and Managing KBS.
	Also attached is Jim Nuzzo's schedule for Educational Services Intel-
	ligent Systems (ESIS).

	Brochures with more detailed course descriptions can be found in the
	AITC library, as well as within the training group.  Please call DTN
	291-8600 to request a brochure sent directly to you.  Another reference
	source is the AIAG::AITC-TECH notes conference where individual courses 
	descriptions/announcements will be posted.







				Program Cycle

		       KNOWLEDGE ENGINEERING CURRICULUM

		         September 1990 - June 1991

	     Class		   Starts:	   	  Ends:	



	      28	         September 5,           October 26, 1990


	      29                 October 24,		December 21, 1990


	      30                 January 2,	        February 22, 1991

	
	      31                 February 20,	        April 12, 1991


              32                 April 24,              June 14, 1991



	

	The Knowledge Engineering Curriculum consists of the following	
	modules:

				VAX/VMS		
		 Overview of AI and Knowledge-Based Systems
	         Introducing and Transferring AI Technology
		         Self-Assessment & Goal Setting
		                  LISP
			 Knowledge Representation
			        VAX OPS5
		    Designing Knowledge-Based Systems
			   Knowledge Acquisition
			       Prototyping
			






	To register, contact Suegene Levin at DTN 291-8600 or send
	mail to her at AIADM::LEVIN.




		






		   EMERGING SOFTWARE TECHNOLOGY TRAINING

			September, 1990 - June, 1991


	Advanced Knowledge Representation	October 9 - November 13,
						3 hours each week

				       
	Knowledge Craft Hands-On Introductory	September 27 - 28,
	Seminar				


	Knowledge Craft				October 1 - 5,


	PAKS Seminar				January 24 - 25, 


	Object Oriented Technology		TBA


	Advanced VAX OPS5			February 11 - 15,


	Blackboard-Based Systems


	Design, Integrating & Testing		March 28 - April 1,
	Knowledge-Based Systems				and
						May 30 - June 3,
	








	
	


	To register, contact Suegene Levin at DTN 291-8600 or send
	mail to her at AIADM::LEVIN.





	





			MANAGING KNOWLEDGE-BASED SYSTEMS

			      Course Schedule 


	Introducing and Transferring AI		September 10 - 13,
	Technology				October 29 - November 1,
						January 7 - 10,
						February 25 - 28
						April 29 - May 2


	Managing Knowledge-Based Systems	November 27 - 30,
						May 20 - 23,






























	To register, contact Suegene Levin at DTN 291-8600 or send
	mail to her at AIADM::LEVIN.










                          SCHEDULE OF EDUCATIONAL SERVICES

                                 INTELLIGENT SYSTEMS
				       
				       (ESIS)
				

          The following courses are offered at the Artificial Intelligence
          Training Center in Marlborough, MA.  In order to register for
          these classes call DTN 291-8893.
          

          Programming in NEXPERT Object       	$1195.00
          5 Day Course
                            Course #EY-8340E-L0

          Offered on the following dates:  	9/17-9/21/90,
          				   	11/12-11/16/90, 
					   	1/14-1/18/91,
                  			   	2/25-2/29/91



          Programming in Advanced NEXPERT   	$1195.00                
	  5 Day Course

          		    Course #EY-A956E-L0
                  
	  Offered on the following dates:  	12/10-12/14/90



          Programming in EPITOOL            	$1195.00                
	  5 Day Course

          		    Course #EY-C265E-L0
                  
	  Offered on the following dates:  	9/10-9/14/90,
      						10/15-10/19/90, 
						12/3-12/7/90



          Programming with MERCURY                $1195.00
          5 Day Course

                             Course #EY-E849E-L0
                  
	  Offered on the following dates:  	8/13-8/17/90,
          					11/5-11/9/90, 
						1/21-1/25/91,           
						3/18-3/22/91





          Programming in VAX OPS5           	$1195.00                
	  5 Day Course

          		      Course #EY-A919E-L0
                  
	  Offered on the following date:  	8/27-8/31/90



          LASER Environment for Object-Oriented
          Programming of C-Based Expert Systems $1195.00
          5 Day Course

          		       Course #EY-F509E-L0
                  
	  Offered on the following dates:  	10/29-11/2/90,
          					12/17-12/21/90, 
						1/28-2/1/91,                    
						3/4-3/8/91



          Programming in KBMS               	$1195.00                
	  5 Day Course

                               Course #EY-F508E-L0
                  
	  Offered on the following dates:  	1/7-1/11/91,
          					3/11-3/15/91



          DEC Trellis Object System &
          Object-Oriented Programming          	$1195.00
          5 Day Course

		               Course #EY-F507E-L0
                  
	  Offered on the following dates:  	2/4-2/8/91,
          3/25-2/29/91



          C++ for  C Programmers            	$1195.00                
	  5 Day Course

          		     Course #EY-F506E-L0
                  
	  Offered on the following date:  	2/18-2/22/91



          C++for Non-C Programmers           	$1195.00
          5 Day Course

          		      Course #EY-F506E-L0
                  
	   Offered on the following date:	2/11-2/15/91
67.110KNOWLEDGE CRAFT TRAININGHERON::ROACHTANSTAAFL !Fri Aug 31 1990 14:5076
Printed by: Pat Roach                                    Document Number: 012900
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     31-Aug-1990 00:12am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: KNOWLEDGE CRAFT TRAINING

********************************************************************************
		EMERGING SOFTWARE TECHNOLOGY TRAINING
********************************************************************************

		           KNOWLEDGE CRAFT		

     DATE:     October 1 - 5, 1990

  LOCATION:    Caribbean Classroom, DLB12
	       295 Donald Lynch Boulevard
	       Marlborough, MA

  DESCRIPTION: Knowledge Craft training is a one-week continuation to the 
	       Two-day Knowledge Craft Hands-On Introductory Seminar.  The
               course focuses primarily on additional functionality of Know-
	       ledge Craft's components.  Topics covered during the training
	       include:

	     o CRL Topics:  contexts, relations, switches, restrictions, user-
	       defined relations, transitivity, inheritance, path restrictions,
	       demons, meta knowledge, dependency maintenance, error handling 
	       and the save manager;

	     o Inferencing strategies: object programming, the agenda mechanism,
	       CRL-OPS and CRL-Prolog;

	     o CRL-OPS: forward chaining, rule-based programming style, conflict
	       resolution, CRL-OPS workcenter and CRL/CRL-OPS integration.

        GOALS: Upon completion of this course, participants will understand
               Knowledge Craft syntax including the:

	     o Knowledge Craft representational and inferencing strategies:
	       CRL (TM), CRL-OPS (TM), CRL-Prolog (TM), Object Programming
	       and Agenda Mechanism;

	     o Knowledge Craft workcenters: CRL workcenter, CRL-OPS workcenter
	       and the CRL-Prolog workcenter; as well as the environments,
	       editors, and utilities; and

	     o Use of window/graphics primitive; and the command system, and 
	       interface task management.

    AUDIENCE &
PREREQUISITES: Significant programming experience with Lisp.  This course is 
	       designed primarily for engineers who will use Knowledge Craft 
	       to design and develop knowledge-based systems.  Attendees 
	       should have prior experience with the Common Lisp language and 
	       the Vax workstation.  Participants should also have a basic 
	       understanding of knowledge-based systems.

	       The Two-Day Knowledge Craft Hands-On Introductory Seminar is 
	       a prerequisite for the one week Knowledge Craft course.

         COST: $2500

LENGTH/FORMAT: 1 weeks, Lecture/Lab

  TO REGISTER: Send mail to AIADM::LEVIN with your 1) name, 2)	node, 3) 
	       organization, 4) manager's name and 5) cost center.
67.111INTRO TO KNOWLEDGE CRAFTHERON::ROACHTANSTAAFL !Fri Aug 31 1990 14:5095
Printed by: Pat Roach                                    Document Number: 012901
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     31-Aug-1990 00:12am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: INTRO TO KNOWLEDGE CRAFT

********************************************************************************
		EMERGING SOFTWARE TECHNOLOGY TRAINING
********************************************************************************

                TWO-DAY KNOWLEDGE CRAFT HANDS-ON INTRODUCTORY SEMINAR

       DATE:    Sept. 27 - 28th, 1990

   LOCATION:	Caribbean Classroom, DLB12-1
		Digital Equipment Corporation
	        295 Donald Lynch Boulevard
		Marlborough, MA 

DESCRIPTION:	The Introductory Seminar is to give the student enough hands-on
	        experience with Knowledge Craft to allow him or her to make an
	        informed decision about using Knowledge Craft for a Knowledge-
		based application. 	

        GOALS:  On completion of the seminar the student will be able to begin 
		work on a prototype of knowledge base and application interface 
		using Knowledge Craft.
	
		Lectures are interspersed with exercises, to allow plenty of 
		hands-on reinforcement of the subject matter.  

		Upon completion of this course, participants will:

		- CRL Basics.  The student will learn how to create knowledge
		representation concepts, and how they are implemented in the
		Carnegie Representation Language (CRL).  The student will learn
		how to create knowledge base objects, how to assign attributes 
		them, and how to modify and delete them.

		- Relations - how objects within a knowledge base can be linked
 		to complete the knowledge representation.  The student learns
	        how to create taxonomic hierarchies, how inheritance occurs
	        using predefined Knowledge Craft relations, and how to create 
	        new, customized relations and use them to link knowledge base
	        objects.

		- CRL Workcenter - the use of the CRL development environment
		The student learns how to use the Graphic Editor to create
		schemata and relations, and how to build networks from them.
		The use of the Save Manager utility is also covered.

		- Object-Oriented Programming - how to use object-oriented 
		programming in Knowledge Craft.  This segment includes a dis-
		cussion of the concepts behind object-oriented programming,
		as well as a comparison with other programming approaches.
		The student solves a sample problem using CRL and object-
		oriented techniques.

		- Introduction of CRL-OPS - an overview of production system
		or rule-based programming.  The student learns about the RETE
		Match Algorithm used by CRL-OPS, how to write production rules
		and how these rules are used in conjunction with objects in
		the knowledge-base.  The CRL-OPS development environment is
		used in solving a sample problem.

		- Rapid Prototyping Methodology (RPM) - an easy approach to 
		developing an application interface.  The student learns how to 
		use this set of utilities to quickly build a graphic interface 
		for application prototype.  The interface includes input and 
		output windows, display windows, buttons (or icons) which acti-
		vate commands using a mouse, and other useful graphics.

      LENGTH:	2 days
	
      FORMAT:   Lecture/Exercises

   PRESENTER:   Carnegie Group, Inc.

	COST:   $1,000/internal price

 TO REGISTER:   Send mail to AIADM::LEVIN.  Please include badge number, DTN,
	        group name, organization name and manager's name.

	THE INTRODUCTORY SEMINAR IS A PREREQUISITE FOR THE ONE-WEEK KNOWLEDGE 
	CRAFT TRAINING COURSE.
67.112ADVANCED K.R. OPENINGSHERON::ROACHTANSTAAFL !Fri Aug 31 1990 14:5175
Printed by: Pat Roach                                    Document Number: 012902
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     31-Aug-1990 00:12am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ADVANCED K.R. OPENINGS

********************************************************************************
	             EMERGING SOFTWARE TECHNOLOGY TRAINING		        
********************************************************************************

 		      ADVANCED KNOWLEDGE REPRESENTATION

	       DATE:    Oct. 9, 16, 23, 30, Nov. 6, 13
	                (3 hours per session)

	   LOCATION:    Archangel Conference Room - DLB12-2
			Digital Equipment Corporation
			295 Donald Lynch Boulevard
			Marlborough, MA  01532

	DESCRIPTION:	This course will survey and discuss major techniques
			and issues in Knowledge Representation using historic/
		        classic and current readings by experts in each tech-
	                nique.  For each representation, the subtleties, 
		        strengths and weakness will be discussed.  For each 
		        meeting, every participant will be expected to write 
			a short critique of the week's readings which will be 
			distributed to all participants.  In addition, at each 
			meeting, a participant will give a short presentation 
			with the instructor, on the major important or contro-
			versial highlights of the readings assigned.  Sources 
			for the readings include: Readings in Artificial Intel-
		        ligence, by Webber & Nelson; Reading in Knowledge Repre-
			sentation, Brachman & Levesque; and Handbook of AI, 
			by Bars & Fergenfaum (eds.).  Participants will be 
			expected to spend time between each session reading the
	                required articles for discussion at the following
		        session.

	       GOALS: 	Upon completion of the course participants will:

			o Understand current Knowledge Representation research 
			  issues.

			o Be aware of alternative Knowledge Representation 
			  techniques, and their appropriateness in applications 
			  development.

			o Be deeply familiar with classical and influential 
			  research papers in the K.R. field.

	    PRESENTER:	Dr. Edwina Rissland, Professor, UMASS, Amherst 
			Computer & Information Systems

      PRE-REQUISITES:   A good understanding of traditional Knowledge 
			Representation	techniques, and an ability to read, 
			discuss and present research papers at the graduate 
			level.

	         COST:  $1200.
	     	      
	       FORMAT: 	Lectures/presentations/discussions/written critiques.

	  TO REGISTER:  Send mail to AIADM::LEVIN.  Please include badge number,
			DTN, group name, organization name and manager's name.
67.113Programming in ANSI CHERON::ROACHTANSTAAFL !Fri Sep 21 1990 18:1675
Printed by: Pat Roach                                    Document Number: 013227
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Sep-1990 09:47pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: Programming in ANSI C

From the NTU Satellite Network:  Live Satellite Broadcast

                                                                                
			ANSI C PROGRAMMING


     COURSE NUMBER:     S9107-01
     DATE:        	October 5, 1990
     TIME:              11 AM - 5 PM 
     LOCATION:          Edgartown Conference Room, LKG2-1/9U                  
     INSTRUCTOR:        Dr. Ira Pohl
     COST:              $350   
--------------------------------------------------------------------------------
     TO REGISTER:  Set Host DELNI; Username: COURSES; Password: not required
                   Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
--------------------------------------------------------------------------------
COURSE OBJECTIVES:

This course will introduce you to programming in ANSI C.  For those 
programmers who have "old" C experience, it will introduce you to ANSI's 
newly introduced features.  Experienced programmers will gain immediate 
access to programming in the new ANSI Standard C.

COURSE DESCRIPTION:

C will be the major programming language of the 1990s for professional 
software developers.  The language was invented by Dennis Ritchie at Bell 
Labs in 1972 and was originally used as a system implementation language to 
implement UNIX.  In the last four years, it has been carefully modified and 
standardized by the ANSI committee to meet modern software requirements.  

By carefully developing working C programs, this course will give you a 
simple and thorough introduction to the programming process in ANSI C.  
Examples of working code will be used to highlight key points.  Sessions 
will show how the language is improved by stronger typing, function 
prototypes, enumerations, void pointers, and other useful additions.  The 
professor will also give advanced details on the use libraries such as 
stdio.h.  From the beginning, you will be introduced to full working 
programs.
                                                                                
INTENDED AUDIENCE:

This course will benefit professional programmers, software designers and 
project leaders desiring familiarity with ANSI C programming.  The course 
assumes experience in another programming language such as FORTRAN, PL/I or 
Pascal.
                                                                                
ABOUT THE INSTRUCTOR:

Dr. Ira Pohl is professor and chairman of the Department of Computer and 
Information Sciences at the University of California, Santa Cruz.  He has 
been a research fellow at the Edinburgh University Machine Intelligence 
Program, a ZWO Professor at the Vrije Universiteit, Amsterdam, and a McKay 
professor at the University of California, Berkeley.  Dr. Pohl is the 
author of seven books.
                                                                                
SPONSOR:

University of Southern California
67.114PROBLEM SOLVING & DECISION MAKING OFFERINGHERON::ROACHTANSTAAFL !Fri Sep 21 1990 19:17104
Printed by: Pat Roach                                    Document Number: 013230
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     20-Sep-1990 11:18pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PROBLEM SOLVING & DECISION MAKING OFFERING

                PROBLEM SOLVING & DECISION MAKING

         COURSE NUMBER:   PROB003
         DATE:            SEPTEMBER 26,-28, 1990
         TIME:            8:30 - 5:00
         LOCATION:        EDGARTOWN C/R LKG2-1
         INSTRUCTOR:      SUZANNE MCMAHON/KEPNER-TREGOE
         COST:            $850
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------



Who should attend:  Managers and professionals  who  want  to  sharpen
their problem solving and decision making skills for improved business
results.

ABSTRACT

This workshop is an intense learning  experience  that  makes  use  of
interesting  case  studies,  role  modeling  and job related concerns.
Upon completion of the workshop, participants will be able to apply  a
proven  rational  management  approach  with  immediate  and  positive
results.

COURSE OUTLINE

Make Decisions:

      o  Make a decision  when  the  choice  between  alternatives  is
         unclear

      o  Prioritize what you really want to accomplish

      o  Evaluate alternatives to generate an overall value  for  each
         one

      o  Make a choice that balances both the benefits and the risks

      o  Sell the decision to others in your organization

Solve Problems:

      o  Avoid panic and scapegoating when a crisis occurs

      o  Collect the facts about a problem efficiently

      o  Focus your search for the  true  cause  by  checking  readily
         available facts

      o  Verify that you have  found  the  true  cause  before  taking
         expensive corrective actions

      o  Deal  with  "people  problems"  as  well   as   machine   and
         organizational problems

Anticipate Potential Problems:

      o  Analyze plans for potential trouble spots

      o  Prevent potential problems from occurring

      o  Develop and implement contingency plans




Manage Complex Issues:

         Breakdown major issues and projects into workable pieces

      o  Prioritize  issues  to  devote  resources  to  most  critical
         concerns

      o  Choose the right analytical tools to handle each issue

      o  Delegate responsibility to accomplish tasks effectively

      o  Case studies that incorporate real-life  decisions,  problems
         and plans

      o  Video tape demonstrations of the management skills

      o  Individual feedback for proper monitoring of your progress

      o  Constant focus on applying the concepts to your job concerns
67.115GRAI METHOD, 10/2, 9:30AMHERON::ROACHTANSTAAFL !Fri Sep 21 1990 19:18119
Printed by: Pat Roach                                    Document Number: 013231
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     21-Sep-1990 00:48am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: GRAI METHOD, 10/2, 9:30AM





	TITLE:  	"Model Classification and the GRAI Methodology"

	SPEAKER:	Dr. Michel Roboam
			Carnegie-Mellon University

	DATE:		Tuesday, 2 October, 1990

	TIME:		9:30 - 11:00AM

	PLACE:		Pacific Room, DLB12

	HOST:		Marsha Hayek

	

	The life cycle of an item of equipment or a manufacturing system
	traditionally includes the five main stages: analysis, design/speci-
	fication, development, implementation and operation.  The difficulties
	in developing exact specifications during the first two phases are 
	now well known. This situation is a direct result of the high degree 
	of automation of new manufacturing systems, the corresponding diffi-
	culties in their analysis and the availability of suitable method-
	ologies which adequately describe the system. It has also been proved 
	that errors in these phases lead to significant costs in subsequent 
	activities.
 
	A large set of modeling tools and methodologies are available to help
	engineers perform these two first phases. The problem we are facing 
	now is finding the right tool for the specific problem we have to 
	study.
 
	In our talk, we will present a classification attempt of all these 
	modeling tools and methodologies based on three different criteria:

	1.  The life cycle of methodologies,
 
	2.  The levels of abstraction covered by the modeling tools, and
 
	3.  The different system taken into account by the methodologies.

 
	The second problem we will address in our talk concern the analysis/
	design of Production Management Systems (PMS).  These systems need 
	methodical procedures because of their complexity and require 	
	competence in many areas, e.g. production management, computer 
	science, networking, organization, computer architecture, knowledge 
	of the technology of the domain, etc. together with knowledge of the 
	implications of a particular PMS for its future users.

	As we will see in our classification, few modeling tools allow a 
	complete description of all the aspects involved in a PMS. Most of 
	these methodologies focus their attention mainly on the information 
	processing (data structures and algorithms). But a PMS is not only 
	software and computers; its main structure is composed of people. 
	In addition, the key points of a PMS concern the decision sequences 
	and the organization (combining both people and computers).

	In early 1980, the GRAI laboratory of the University of Bordeaux I 
	(France) focussed its attention on the problem of analysing and 
	designing production management systems. They defined the first 
	basis of a methodology dedicated to the modeling of PMS in 1984 
	through two PhD theses ([Doumeingts.84] and [Breuil.84]). Then, 
	through a loop "application-criticism-research-improvment", 
	they created the GRAI method ([Dufresne.86] and [Roboam.88]).
 
	Today, this methodology is used by several large engineering 
	companies such as CAP-GEMINI-SOGETI, SEMA-METRA, SGN, .... It 
	has been selected by several European ESPRIT projects as the 
	modeling tool for the specification of the different systems 
	and architectures developed.
 
 
	Dr. Michel Roboam is a Visiting Research Scientist in the Center for 
	Integrated Manufacturing Decision Systems of the Robotics Institute at 
	Carnegie Mellon University. He received a B.S. in Mathematics in June 
	1983 and a M.S. in Electronic, Electrotechnic and Automatic in June 
	1985 from the University of Bordeaux I (France).  In September 1985, 
	he joined the GRAI Laboratory of the University of Bordeaux I as a 
	Ph.D. student. 
	
	He received his D.E.A in June 1986 and his Ph.D. in CIM in July 1988 
	from the University of Bordeaux I. He was responsible for the GRAI 
	Laboratory of the European ESPRIT Project 418 (Open CAM System) for 
	two years. In February 1989, he joined the Center for Integrated 
	Manufacturing Decision Systems. 
	
	Dr Roboam's research interests are in production management, 
	methodologies, knowledge representation and artificial intelligence. 
	His current work is focused on decentralized architectures and on 
	the utilization of artificial intelligence techniques to support the 
	integration of manufacturing systems. 

	Dr Roboam is currently involved in the EMN (Enterprise Management 
	Network) project, whose aim is to develop an architecture able to 
	support the integration of manufacturing activities.  He is working 
	on the design of an Intelligent Network Architecture.  He is the 
	designer and developer of the decentralized communication system of 
	this architecture.
 
67.116PROGRAMMING IN MERCURYHERON::ROACHTANSTAAFL !Tue Oct 30 1990 14:08122
Printed by: Pat Roach                                    Document Number: 013641
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     30-Oct-1990 11:04am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PROGRAMMING IN MERCURY


         		PROGRAMMING IN MERCURY KBE


        COURSE DATES:     November 5-9, 1990

	COURSE LOCATION:  Artic Room, DLB12-1

**-->   TO REGISTER:      Call the ESIS group at 291-8893, or 
	                  send EMAIL to ESIS::WILSON. 



        DESCRIPTION:
	 
	Mercury Knowledge-Based Environment (KBE), a product of Artificial 
	Intelligence Technologies, Inc. (AIT), is a high-performance tool for 
	developing and delivering intelligent, expert-system-based 
	applications. It provides a robust set of features in a seamlessly 
	integrated, extensible environment, such as:

	o Knowledge representation paradigms.
	o Object-oriented programming.
	o Rule-based programming.
	o Productivity enhancement tools.

	Mercury KBE provides:

	o A seamless integration of object, database, and rule-based concepts.
	o Consistent knowledge representation.
	o The ability to prototype and deliver applications in the same 
	  environment.

	Mercury KBE is layered on the Common LISP programming environment, 
	thus providing an extension to this environment, one that is:

	o Extremely efficient.
	o With minimal garbage.
	o Using minimal resource consumption.


	COURSE LAYOUT:

	This lecture/lab Training Course is designed to let you become familiar 
	with using the Mercury KBE system. Labs will provide students with a 
	hands-on experience working with Mercury on VAX Workstations, while 
	lectures provide explanations of concepts and functionality of the tool.
	
	In general, we will be discussing:

	o Basic Mercury KBE and object concepts.
	o Introductions to all Mercury KBE components.
	o Advanced uses of Mercury KBE.

	Here's a brief description of the major Mercury KBE components covered 
	in the course:

	o Mercury Object System

	The Mercury Object System (MOS) is the object-oriented component of
	the Mercury KBE system.  MOS implements a full-featured extension to 
	Common LISP that is as efficient as possible.  In general, MOS is 
	compatible with CLOS, the Common LISP Object System. Occasional
	modifications have been made that increase efficiency.

	o Mercury Production Engine

	The Mercury Production Engine allows you to implement highly efficient,
	rule-based code.  This facility reasons over MOS objects.  The Mercury
	Production Engine operates in a RETE-like manner with regard to pattern
	matching, rule firing, and the general syntax of the language.

	o Mercury Structured Query Language

	The Mercury Structured Query Language (MSQL) allows you to access 
	database information in a Mercury KBE application.  Currently, Mercury 
	KBE supports interfaces with the Rdb and Oracle databases. Support for 
	other widely-used relational databases is planned for future Mercury 
	KBE releases.

	o MOS-CLASS Metaclass

	The MOS-CLASS metaclass is the interface among all of the components 
	of Mercury KBE for the following areas:
 
	   o Support for various types of user interfaces.
	   o Provides an MSQL/Object interface with allows instance objects to 
	     be read from, stored in, and deleted from databases.
	   o Provides a Production Engine/Object interface which allows 
	     inferencing over MOS-CLASS objects.
	   o Support for the definition of relations between object classes.


        WHO WILL BENEFIT FROM THIS COURSE:

        Programmers and Knowledge Engineers who need to develop an
        expertise in a medium to large scale Expert System development
        tool.
         

        PREREQUISITES:

	Digital Educational Services' "Symbolic Programming in VAX Lisp - I" 
	or the equivalent. Previous experience with expert systems is not 
	necessary but may be helpful.
        
67.117PROGRAMMING IN MERCURYHERON::ROACHTANSTAAFL !Mon Nov 19 1990 11:58122
Printed by: Pat Roach                                    Document Number: 013905
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     16-Nov-1990 08:49pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: PROGRAMMING IN MERCURY


         		PROGRAMMING IN MERCURY KBE


        COURSE DATES:     December 3-7, 1990

	COURSE LOCATION:  Arctic Room, DLB12-1

**-->   TO REGISTER:      Call the ESIS group at 291-8893, or 
	                  send EMAIL to ESIS::WILSON. 



        DESCRIPTION:
	 
	Mercury Knowledge-Based Environment (KBE), a product of Artificial 
	Intelligence Technologies, Inc. (AIT), is a high-performance tool for 
	developing and delivering intelligent, expert-system-based 
	applications. It provides a robust set of features in a seamlessly 
	integrated, extensible environment, such as:

	o Knowledge representation paradigms.
	o Object-oriented programming.
	o Rule-based programming.
	o Productivity enhancement tools.

	Mercury KBE provides:

	o A seamless integration of object, database, and rule-based concepts.
	o Consistent knowledge representation.
	o The ability to prototype and deliver applications in the same 
	  environment.

	Mercury KBE is layered on the Common LISP programming environment, 
	thus providing an extension to this environment, one that is:

	o Extremely efficient.
	o With minimal garbage.
	o Using minimal resource consumption.


	COURSE LAYOUT:

	This lecture/lab Training Course is designed to let you become familiar 
	with using the Mercury KBE system. Labs will provide students with a 
	hands-on experience working with Mercury on VAX Workstations, while 
	lectures provide explanations of concepts and functionality of the tool.
	
	In general, we will be discussing:

	o Basic Mercury KBE and object concepts.
	o Introductions to all Mercury KBE components.
	o Advanced uses of Mercury KBE.

	Here's a brief description of the major Mercury KBE components covered 
	in the course:

	o Mercury Object System

	The Mercury Object System (MOS) is the object-oriented component of
	the Mercury KBE system.  MOS implements a full-featured extension to 
	Common LISP that is as efficient as possible.  In general, MOS is 
	compatible with CLOS, the Common LISP Object System. Occasional
	modifications have been made that increase efficiency.

	o Mercury Production Engine

	The Mercury Production Engine allows you to implement highly efficient,
	rule-based code.  This facility reasons over MOS objects.  The Mercury
	Production Engine operates in a RETE-like manner with regard to pattern
	matching, rule firing, and the general syntax of the language.

	o Mercury Structured Query Language

	The Mercury Structured Query Language (MSQL) allows you to access 
	database information in a Mercury KBE application.  Currently, Mercury 
	KBE supports interfaces with the Rdb and Oracle databases. Support for 
	other widely-used relational databases is planned for future Mercury 
	KBE releases.

	o MOS-CLASS Metaclass

	The MOS-CLASS metaclass is the interface among all of the components 
	of Mercury KBE for the following areas:
 
	   o Support for various types of user interfaces.
	   o Provides an MSQL/Object interface with allows instance objects to 
	     be read from, stored in, and deleted from databases.
	   o Provides a Production Engine/Object interface which allows 
	     inferencing over MOS-CLASS objects.
	   o Support for the definition of relations between object classes.


        WHO WILL BENEFIT FROM THIS COURSE:

        Programmers and Knowledge Engineers who need to develop an
        expertise in a medium to large scale Expert System development
        tool.
         

        PREREQUISITES:

	Digital Educational Services' "Symbolic Programming in VAX Lisp - I" 
	or the equivalent. Previous experience with expert systems is not 
	necessary but may be helpful.
        
67.118ADVANCED COURSE - K.R.HERON::ROACHTANSTAAFL !Tue Nov 20 1990 09:5078
Printed by: Pat Roach                                    Document Number: 013927
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Nov-1990 09:03pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: ADVANCED COURSE - K.R.

	  TO REGISTER:  Send mail to AIADM::LEVIN.  Please include badge number,
			DTN, group name, organization name and manager's name.

********************************************************************************
	             EMERGING SOFTWARE TECHNOLOGY TRAINING		        
********************************************************************************

 		      ADVANCED KNOWLEDGE REPRESENTATION

	       DATE:    April 16, 23, 30, May 7, 14, and 21
	                (3 hours per session; 9:30 - 12:30 p.m.)

	   LOCATION:    DLB12-2
			Digital Equipment Corporation
			295 Donald Lynch Boulevard
			Marlborough, MA  01532

	DESCRIPTION:	This course will survey and discuss major techniques
			and issues in Knowledge Representation using historic/
		        classic and current readings by experts in each tech-
	                nique.  For each representation, the subtleties, 
		        strengths and weakness will be discussed.  For each 
		        meeting, every participant will be expected to write 
			a short critique of the week's readings which will be 
			distributed to all participants.  In addition, at each 
			meeting, a participant will give a short presentation 
			with the instructor, on the major important or contro-
			versial highlights of the readings assigned.  Sources 
			for the readings include: Readings in Artificial Intel-
		        ligence, by Webber & Nelson; Reading in Knowledge Repre-
			sentation, Brachman & Levesque; and Handbook of AI, 
			by Bars & Fergenfaum (eds.).  Participants will be 
			expected to spend time between each session reading the
	                required articles for discussion at the following
		        session.

	       GOALS: 	Upon completion of the course participants will:

			o Understand current Knowledge Representation research 
			  issues.

			o Be aware of alternative Knowledge Representation 
			  techniques, and their appropriateness in applications 
			  development.

			o Be deeply familiar with classical and influential 
			  research papers in the K.R. field.

	    PRESENTER:	Dr. Edwina Rissland, Professor, UMASS, Amherst 
			Computer & Information Systems

      PRE-REQUISITES:   A good understanding of traditional Knowledge 
			Representation	techniques, and an ability to read, 
			discuss and present research papers at the graduate 
			level.

	         COST:  $1200 for internal people.
	     	      
	       FORMAT: 	Lectures/presentations/discussions/written critiques.

	  TO REGISTER:  Send mail to AIADM::LEVIN.  Please include badge number,
			DTN, group name, organization name and manager's name.
67.119KNOW. CRAFT-PAKS SEMINARHERON::ROACHTANSTAAFL !Tue Nov 20 1990 09:5188
Printed by: Pat Roach                                    Document Number: 013929
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     19-Nov-1990 09:19pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: KNOW. CRAFT-PAKS SEMINAR


  TO REGISTER:  Send mail to AIADM::LEVIN with your name, node, organization,
		manager's name and cost center.

********************************************************************************
   	        EMERGING SOFTWARE TECHNOLOGY TRAINING
********************************************************************************

                KNOWLEDGE CRAFT TRAINING - PAKS SEMINAR
		(SIMULATION AND RELATED MODULES)

       DATE:    Jan. 24-25, 1991

   LOCATION:    Caribbean Classroom, DLB12-1
	        295 Donald Lynch Boulevard
	        Marlborough, MA

DESCRIPTION:	The PAKS course will provide the student with a thorough under-
		standing of how to use the "pak" modules of Knowledge Craft:
		OOPpak, Statpak, Simpak, Graphpak and Timepak.  The student
		will get a chance to use the software during exercise and lab
		sessions.  On completion of the course the student will know 
		how to incorporate the functionality of the "pak" modules into 
		an application.

      GOALS:    The following topics are covered:

		- OOPpak - the object-oriented functionality which is used to
		implement the rest of the "pak" modules.  Basic concepts of
		object-oriented programming are reviewed.  The student learns
	        to create objects, define and attach methods to the objects,
		and send messages to the objects.

		- Simpak - the discrete event simulation module.  The student
		learns to create a simulation, to create and manage events
		associated with the simulation, and to execute the simulation.

		- Statpak - Knowledge Craft's statistical functionality.  The
		various statistical distributions provided with Statpak and how
		they can be used in a simulation application are covered.  The
		student learns to use data collectors and instruments for 
		gathering and measuring statistical information generated within
		a Knowledge Craft application.

		- Graphpak - business graphs and gauges.  The student learns to
		create and use these graphic objects, both for display of
		information and as active graphics so the end user of the appli-
		cation can enter data.

		- Timepak - temporal representation.  The student learns to 
		create a representation of absolute time and of time intervals
		using the tools available with Timepak.  Also covered is the
		representation of relative temporal concepts and of relations
		among objects based on these temporal concepts.

PREREQUISITES:  A working knowledge of the basic Knowledge Craft functionality
		will ensure that the student is able to make the most of this
		seminar.  The student should be able to create and manipulate
		schemata, make use of relations, create and use canvases, 
		windows and viewports, and create and manipulate graphic
		objects.

   PRESENTERS:  Carnegie Group, Inc.

       LENGTH:	2 days

       FORMAT:  Lecture/Exercises
  
 	 COST:  $1,000/internal charge

  TO REGISTER:  Send mail to AIADM::LEVIN with your name, node, organization,
		manager's name and cost center.
67.120REMINDER-11/28-S/W PATENTSHERON::ROACHTANSTAAFL !Tue Nov 27 1990 21:06109
Printed by: Pat Roach                                    Document Number: 014064
--------------------------------------------------------------------------------

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     27-Nov-1990 08:44pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@MRGATE@HUGHI
                                        Dept:      
                                        Tel No:    

TO:  ROACH@A1NSTC


Subject: REMINDER-11/28-S/W PATENTS







	TITLE:		"Software Patents And The Digital
		 	Patent Process"

	NOTE:           THIS SEMINAR WILL **NOT** BE VIDEOTAPED
			---------------------------------------
					     	
	SPEAKERS:   ->  JOHN MESAROS, Manager
			Patent Services Law Group
			
		    ->  KERRI SENZAPAURA, Administrator
			Corporate Patent Incentive Award Program

			Note:  Introduction of speakers
			by Vin Ranucci, Digital Patent Attorney  	    
	
	DATE:		Wednesday, November 28, 1990

	TIME:		9:30-11:00AM

	PLACE:		LEOPARD CONFERENCE ROOM
			DLB5-3/E10


        Did you know that Digital gives monetary ($$$) awards to 
	employees who file and are awarded patents?

	There is significant business value in having patents. 
	And, Digital is very interested in increasing its patent 
	stable. 
	
    	Find out more about the monetary awards, more about the 
	value of patents and why DEC needs them, a bit about the 
	process of filing for a patent, and even some of the 
	amazing things that have already been patented (you 
	wouldn't believe!...)

        When you were a kid, did you want to be an inventor?  What
        better way to prove you're an inventor than to get a patent
        on something you've invented!

    	Not too long ago, software inventions couldn't be patented.  
	Now, it is well established that they can be, and AI, in 
	particular, already has a number of patent applications in 
	the pipeline.  We need MORE!





	NOTE:     THIS SEMINAR WILL **NOT** BE VIDEOTAPED
                  ---------------------------------------








	TITLE:		"Software Patents And The Digital
		 	Patent Process"
	Page Two

	    


  	About the speakers...

    	JOHN Mesaros is the manager for the Patent Services Law Group.  
	He received his BSEE from Carnegie Tech in 1958 and his J.D. 
	from the University of Pittsburgh in 1964.  He started in the 
	patent business in 1960.  He has had 15 years of experience 
	with corporations and 15 years in private patent practice.  
	JOHN joined Digital in January of 1989.

	Kerri Senzapaura is the Corporate Patent Incentive Awards 
	Administrator for the Incentive Award Program.  Kerri 
	received her B.S. in Marketing from the University of 
	Lowell in 1987.  She has been the program's administrator 
	for one year, and is charged with updating its policy and 
	procedures under the direction of JOHN Mesaros and the 
	Corporate Intellectual Property Protection Committee.



	NOTE:     THIS SEMINAR WILL **NOT** BE VIDEOTAPED
                  ---------------------------------------
67.121New notes for conference EXPERTULYSSE::ROACHTANSTAAFL !Wed Jan 30 1991 10:4793
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     30-Jan-1991 10:32am CET
                                        From:     ROACH
                                                  ROACH@YIPPEE@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: New notes for conference EXPERT

------------------------AI Courses at DLB12-------------------------------------
206.1   LISPER::WILSON        -< Programming in KBMS: Feb. 11 - 15, 1991 >-

		Notes start on next page.


               <<< BROKE::STAR01:[NOTES$LIBRARY]EXPERT.NOTE;1 >>>
                              -< Expert Systems >-
================================================================================
Note 206.1                     AI Courses at DLB12                        1 of 1
LISPER::WILSON                                       59 lines  25-JAN-1991 15:48
                  -< Programming in KBMS: Feb. 11 - 15, 1991 >-
--------------------------------------------------------------------------------

Programming in KBMS 
Course Number: EY-F508E-L0
Length: 5 days

<OverView>

KBMS (Knowledge Based Management System) is an expert system tool developed by
AICorp. KBMS offers high-level, seamless integration with different database
management systems and the ability to be called from existimg conventional
application programs. The KBMS architecture makes it highly flexible for use in
developing a wide variety of applications and for enhancing existing product
applications. The course Programming in KBMS is a five day course that covers
the use and features of KBMS knowledge base development and use.

<Objectives>
The course consists of lectures and labs.  During the lectures concepts
and features of KBMS are covered.  Labs will provide the students
with a hands on experience working with KBMS on VAX workstations.
The goals of the course are to enable the participants to
  -  develop an expert system using KBMS components and inferencing strategies
  -  retrieve database information using the natural language facilities in
     KBMS
  -  effectively use KBMS features for enhancing fluency and rule definition
  -  efficiently access internal and external object occurrences
  -  evaluate expert system performance and tune KBMS applications  

<Prerequisites>

It is assumed that students will come into the course with knowledge of at
least one programming language. Previous experience with expert systems is not
necessary but may be helpful. 

<Topics>
  *  KBMS Architecture
     -  Knowledge Base
     -  Inference Engine
     -  Developer Interface
     -  Database Interface
     -  KBMS Environments
  * KBMS Developer Interface
     -  KBMS Editor
     -  Compilation
     -  Running a session
     -  Utilities and Actions
  * KBMS Knowledge Representations
     -  Objects and Attributes
     -  Packets
     -  Rules
  * Strategic and Tactic Inference in KBMS
     -  Forward Chaining
     -  Backward Chaining
  * KBMS Natural Language Query System
     -  Control Access and Retrieval
     -  Session Commands and Modifiers
     -  Object and Record Management
  * Knowledge Base Maintenance in KBMS
     -  KBMS Development Cycle
     -  Performance and Tuning
		<<< End of extracted notes >>>

Extraction performed by AIOLI::ROACH at 10:23 on 30-Jan-91 (Wed)
 using AIOLI::DISK2:[AI$USER.ROACH.NOTES]NOTES.COM;1 -- v37
   with switches "/name=<ROACH_NOTES_WEEKLY> /class=<general_ai> /again_at=<today+7-4:00> /Not_Now=<No>"

67.122Architecting O-O Software Course OfferingULYSSE::ROACHTANSTAAFL !Mon Feb 04 1991 13:0970
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     01-Feb-1991 00:51am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Architecting O-O Software Course Offering

                      ARCHITECTING OBJECT ORIENTED SOFTWARE

         COURSE NUMBER:  L913501
         DATE:           MARCH 4-8, 1991
         TIME:           8:30 - 5:00
         LOCATION:       T&N TRAINING ROOM TAY2-1
         INSTRUCTOR:     TBA
         COST:           $1200
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
Course Goals:
   o  Understand the fundamentals of object oriented technology from a 
      product lifecycle point of view
   o  Understand the paradigm shift and cost/benefits of the object 
      oriented process

Description:
This five day course gives a mid-level, horizontal slice through 
object oriented technology.  The course was developed in order to 
give experienced software professionals a detailed exposure to
object oriented analysis, design, and programming followed by a
mapping of the implementation into the C++ language.  The course
consolidates concepts with desk exercises for analysis and design,
and hands on labs for the C++ implementation.  A common example
thread is used to consolidate the steps of the process.

Topics:
   o  Introduction of Common Example Thread
   o  Object Oriented Analysis - Grammatical Dissection
   o  Conceptual Object Oriented Design - CRC and collaborations
   o  Architecting an Object Oriented Model
      -  Language independent modelling tools
      -  C++ features which support the object oriented model
         o  Costs
         o  Benefits
         o  Blackholes
  o  Object Oriented Programming in C++
     -  Synthesis of modelling tools:
        o  Program organization
        o  Class hierarchies
        o  Structure hierarchies
        o  Control flow
     -  Efficient Implementation of Object Oriented Designs in C++
  o  Testing and Performance Evaluation

Audience:
  o  Software product developers

Prerequisites:
  o  Experience in developing software products
  o  Familiarity with analysis, design, implementation issues
     (i.e., waterfall model)

67.123C for Experienced Programmers Course OfferingULYSSE::ROACHTANSTAAFL !Mon Feb 04 1991 13:1141
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     01-Feb-1991 00:51am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: C for Experienced Programmers Course Offering

                          C FOR EXPERIENCED PROGRAMMERS

         COURSE NUMBER:  L913701
         DATE:           MARCH 18-22, 1991
         TIME:           8:30 - 5:00
         LOCATION:       T&N TRAINING ROOM TAY2-1
         INSTRUCTOR:     HARVEY DEITEL/INFORMATIVE STAGES
         COST:           $1000
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
Brief Description:

This course is an intense five-day, lecture-only format.
It is designed for programmers who have used another 
language intensively, but who have little or no C experience.
For those learning C for the first time.

Prerequisites:
   1.  Considerable programming experience in another programming
       language

   2.  General familiarity with program development in a VAX/VMS
       environment

67.124C Laboratory Workshop OfferingULYSSE::ROACHTANSTAAFL !Mon Feb 04 1991 13:1945
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     01-Feb-1991 02:01am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: C Laboratory Workshop Offering

                      C LABORATORY WORKSHOP

         COURSE NUMBER: L913901
         DATE:          MARCH 11-15, 1991
         TIME:          8:30 - 5:00
         LOCATION:      T&N TRAINING ROOM TAY2-1
         INSTRUCTOR:    HARVEY DEITEL/INFORMATIVE STAGES
         COST:          $1000
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
Brief Description:

This course is an intense, customized, five-day laboratory experience
in C. An individual work plan is prepared for each student based on 
his or her C experience.  Students may work individually or in groups.
enrollment is strictly limited to offer the most effective laboratory
environment.  Programming problems and exercises are presented from
Dr. Deitel's new C book (to be published by Prentice Hall, 1991).
The course is personally supervised by Dr. Deitel who will work with
each student to develop effective C program design, build, and test
skills.

Prerequisite:

This course is appropriate for students with any level of C experience
including at least an introductory course in C.  The student must be
generally familiar with program development in a VAX/VMS environment.

67.125Seminar Offering @ Cambridge, "Navlab and AMBLER",ULYSSE::ROACHTANSTAAFL !Mon Feb 04 1991 13:2070
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     01-Feb-1991 03:11am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Seminar Offering @ Cambridge, "Navlab and AMBLER", 2/25, 3:15PM

                      Digital Equipment Corporation
                         Cambridge Research Lab
                                Seminar
 
                       Monday, February 25, 1991
                 3:15pm - 4:15pm, refreshments at 3:00pm
 
                  "Navigation at CMU: Navlab and AMBLER"
 
                               Chuck Thorpe
              Robotics Institute, Carnegie Mellon University
 
The two primary testbeds for outdoor navigation at CMU are the
Navlab, a computer-controlled Chevy van, and the AMBLER, a walking
machine for planetary exploration.  Research on the Navlab includes
color vision, 3-D data interpretation, path planning, and map
building.  There are three different approaches for following roads:
pattern classification, feature tracking controlled by an AI system,
and neural nets.  The most ambitious system to date drives the
Navlab around an unmodified suburban neighborhood, using neural
nets for road following, 3-D vision for landmark recognition,
inertial navigation, and annotated maps.  Our latest video tape
shows autonomous Coca-Cola delivery.
 
The AMBLER has a novel 6-legged design, with the three left legs and
three right legs mounted on common shoulder shafts.  This allows for
very efficient motion in rugged terrain.  It also brings up
interesting issues in planning, including the "conservative support
polygon", a concept that guarantees that even if any one leg fails,
the vehicle will not tip over.
 
------------------------------------------------------------------------------
 
Chuck Thorpe is a Senior Research Scientist at the Robotics
Institute of CMU.  His PhD is from the Computer Science Department
of CMU in 1984.  His interests are in vision, navigation, and system
building for mobile robots working outdoors.  He is principal
investigator of the Navlab project, and also works on robots for
space and underwater applications.
 
------------------------------------------------------------------------------
 
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and
Broadway.  The address is:
 
     Digital Equipment Corporation
     Cambridge Research Lab
     One Kendall Square, Bldg. 700
     Cambridge, MA 02139
 
For questions regarding the seminar, directions, or parking, or to add 
your name to the mailing list, call DTN: 259-6601 or send email to 
crl::tss

67.126REMINDER - Here's an opportunity to find out aboutULYSSE::ROACHTANSTAAFL !Mon Feb 04 1991 18:06114
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     04-Feb-1991 05:18pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: REMINDER - Here's an opportunity to find out about a project in the AITC-- Please join us at 10AM on 2/5, Pacific Room, "Design-With Features System"






	TITLE:  	"Designing Mechanical Components With Features: 
			Representing the Form and Intent of In-Progress 
			Designs for Automated Modification and Evaluation"

	SPEAKER: 	ERIC NIELSEN
			Senior Software Engineer
                  	ISTG/Concurrent Engineering Group

	DATE:    	5 February, 1991

	TIME:    	10:00 - 11:30 AM

	PLACE:		Pacific Room, DLB12-1

	HOST:    	GARRY WOODRUFF 
			DASH Program Manager
         		ISTG/Concurrent Engineering Group
			AI Technology Center
 
	A feature-based mechanical design system for thin walled components 
	will be described.  It is a Design-With Features System (DWFS). And,
	it has been developed and implemented in an experimental computer 
	program. 

	This seminar will provide an example of how a designer interacts
	with the DWFS, as well as some of its underpinnings. 

	"Features", in this system, are defined as "form plus intent". They 
	are used for entry, modification, and evaluation of the design. 

	To create this system, advances were required in the representations 
	of form and intent, as well as the ability to propagate modifications 
	to the geometry of the design, based on intent. 

	This research has advanced these three areas, and has coupled them 
	together, to create the core of the DWFS.  These key results provide 
	several new and promising ideas, which may form the basis for more 
	intelligent computer-based mechanical design processes in the future.
 
	Though the DWFS may be compatible with current design procedures,
	its use is expected to promote a different process - a process,
	which will enable the production of a more manufacturable design,
	more quickly. 







	DWFS Seminar, 2/5/91
	Page Two




	The process would take advantage of the DWFS's product modelling, 
	evaluation, and modification capabilities. The important character-
	istics of the DWFS are: 

		o composing the form of the design entirely as a 
	          configuration level of features
		o flexibly representing the geometric intent of 
	          the design, and, thus, its level of commitment
		o automatic, interactive (i.e., fast) propagation 
		  of geometry modifications consistent with intent, 
		  whether over-, under-, or fully-"constrained"
		o supporting user-defined features, while limiting 
	          designers in some other ways
		o ability to support interactive knowledge-based 
	          (i.e., feature-based) evaluations of designs 
	          (including user-defined feature-forms)
		o encouraging a least commitment design process

	These characteristics are all implemented in a manageable, 
	extensible, and computationally efficient way.
 	
	This DWFS is shown to be a viable candidate as the basis of 
	future CAD/CAE systems.

	==============================================================

	Eric recently completed his Ph.D. in Mechanical Engineering from 
	U. Mass. at Amherst, under Prof. Jack Dixon. His degree work 
	included some formal computer and information science training.

	Between getting his Bachelor's degree from Cornell University and 
	his Master's degree from U. Mass., Eric spent four years in
	Quality Assurance and Test Engineering for both GTE and Raytheon,
	working on military communications equipment. 

	He is currently a member of the DASH design advisor shell project
	at Digital's AI Technology Center in Marlborough. 

67.127A few slots open, and limited time to sign up - BlaULYSSE::ROACHTANSTAAFL !Fri Feb 08 1991 12:0987
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     06-Feb-1991 10:51pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: A few slots open, and limited time to sign up - Blackboard-based AI Sys.

********************************************************************************
	             EMERGING SOFTWARE TECHNOLOGY TRAINING		        
********************************************************************************

        COURSE NAME:	BLACKBOARD-BASED AI SYSTEMS

================================================================================
	TO REGISTER:    Send mail to AIADM::LEVIN.  Please include badge number,
			DTN, group name, organization name and manager's name.
================================================================================

	      DATES:	March 18-20, 1991

	      PRICE:    $1,500.

	   LOCATION:    DLB12-2
			Digital Equipment Corporation
			295 Donald Lynch Boulevard
			Marlborough, MA  01532

DESCRIPTION:	The blackboard paradigm is a powerful and flexible problem-
	        solving architecture.  This course provides a detailed 
                examination of the blackboard paradigm and laboratory 
                experience building a blackboard-based AI application.  
                Emphasis is placed on the advantages/disadvantages of 
		the blackboard approach and on characteristics of applications 
	        that make them well-suited to blackboard technology. Finally, 
		the blackboard paradigm provides a number of problem-solving 
                control opportunities, and we consider recent developments 
		in blackboard control and meta-level control mechanism.

GOALS:		Upon completion of the course, participants will have:

		o A working knowledge of how blackboard-based AI systems 
                  function and an understanding of what types of applications 
                  are appropriate for blackboard architectures.

		o A historical perspective on the development of blackboard 
                  technology over the last decade and a view of where research 
                  on blackboard approaches is headed in the coming years.

		o An understanding of the efficiency and control issues that 
                  must be addressed in developing a high-performance 
                  blackboard-based application.

		o Had the opportunity to build a small blackboard-based 
                  scheduling application using the Generic Blackboard 
                  Development System (GBB).







********************************************************************************
	             EMERGING SOFTWARE TECHNOLOGY TRAINING		        
********************************************************************************

        COURSE NAME:	BLACKBOARD-BASED AI SYSTEMS
********************************************************************************

PREREQUISITE:	General AI knowledge, programming ability in LISP or OPS5, and 
                knowledge of an editor

LENGTH:		3 Days

FORMAT:		Lecture/lab

PRESENTER:	Dr. Daniel Corkill
		Dr. Victor Lesser, UMASS Amherst

67.128Panel discussion on DEC Business in Asia, today atULYSSE::ROACHTANSTAAFL !Mon Feb 25 1991 18:31103
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     25-Feb-1991 06:18pm CET
                                        From:     AITCVOD
                                                  AITCVOD@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Panel discussion on DEC Business in Asia, today at 11:30am.

Event:	  DEC Business in Asia and Asian Employees at DEC.
Sponsors: AITC Valuing of Diversity Task Force.
Date:     Feb 25, 1991 (Monday)
Time:     11:30am to 1pm
Place:    LM02 (Marlboro), Room 144

Description:
------------
Realizing the importance of DEC business in Asia and contributions of Asian 
employees in DEC, AITC VOD Tasks Force at DLB/LMO cluster has
organized an panel discussion on DEC business in Asia. Since some of
the panelists are of Asian origins, they will also talk about their
unique working experiences at DEC because of their cultural background. 

It is an excellent opportunity to know more about DEC Business in Asia
and share experiences with employees with cultural diversity.

Some Asian ethnic food may be served at the LMO2 Cafe on that day.

Here's a description of the panelists in alphabetical order. 

Ram Josyula
-----------
Principal Engineer in the digital Business Fellowship Program in
Artificial Intelligence. Works with DEC's important clients in selecting
AI Fellows who come to DEC to a six month Fellowship and mentors them in
learning and transferring the technologies to their respective organizations.
In this capacity, Ram has been working with customers from almost twenty
countries.  He came to USA from India in 1980 and been working as an 
AI consultant since 1984.  He joined Digital AI Technology Center in 1987.

Tony Leung
----------
ALPHA PROGRAM MARKETING MANAGER, RESPONSIBLE FOR THE INTRODUCTION 
PLANNING ACTIVITIES FOR THE ALPHA PROGRAM, FOCUS ON ANNOUNCEMENT EVENTS 
PLANNING, MIGRATION REQUIREMENTS, ETC.

Experience in doing business in Asia:
8 YEARS.  2 YEARS IN THE HONG KONG BRANCH OFFICE, 6 YEARS IN REGIONAL 
OFFICE.

HK BORN, EDUCATED IN US (HIGH SCHOOL, COLLEGE), CURRENTLY ON 3 YEAR 
ASSIGNMENT.  WITH DEC FOR 8 YEARS, FIRST JOB.  IN THE US FOR 6 MONTHS.


Neil Pundit
-----------
SW Engrg Group Mgr, R&D in Advanced Service Delivery Systems 

Exp in doing business in Asia: 4 years as a lecturer, exposure to
Indian space programs, employing DE(I)L engrs

Neil came to US in 1965 for higher studies; graduate 
degrees in EE; major technical contribution in guidance and control of
space vehicles; US space programs 10 years with NASA, JPL/CALTECH, TRW;
Dir of Tech activities at IEEE Hq for 3 years; with DEC since 1980;
founding mgr of the second oldest AI group in DEC.

Wayne Purnell
-------------

Functional Manager for human resource development for GIA M&E.
Supports the GIA HQ groups and plants with respect to HRD strategy, programs, 
processes, etc.  Wayne has been working in 
GIA for 3 1/2 years and has spent a fair amount of time in and working with
the Asian plants, especially Hong Kong, PRC, Taiwan, and Singapore.  
Wayne worked previously in Europe (with DEC) for a few years, and before 
that in the U.S. both in Mfg and the Field.  
Wayne has also been involved with cross-cultural training efforts.  



Simon Szeto
-----------
Software Engineer Manager.  Manager of the ZKO site of ISEDA (Inter-
national Systems Engineering, Development and Architecture).  ZKO is one
of seven ISEDA sites around the world and the only site in the US.

Simon  managed the ABSS (Asian Base Systems Software) group in Hong Kong from
December, 1987 to June, 1990.  ABSS became the Asian component of ISE
which was formed in July, 1990.  The mission of ABSS was to produce
Asian versions of Digital's strategic software products.  This is still
ISE's mission but ISE has a global scope, not just Asian.

Simon was born in China, but other than that technicality he was essentially a
native of Hong Kong.  Came to the US in 1961.  Graduated in 1965 with a
B.S. degree.  Lived in New England since 1968.  With DEC 15 years.

67.129FWD: AI SEMINARULYSSE::ROACHTANSTAAFL !Tue Mar 26 1991 11:51367
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     04-Mar-1991 05:55pm CET
                                        From:     BEANE
                                                  BEANE@BIGRED@MRGATE@DPD05@DPD
                                        Dept:      
                                        Tel No:    

TO: See Below

Subject: FWD: AI SEMINAR

<<forwarding deleted>>

From:	NAME: Fred Mariani                  
	FUNC: ESG/EIAM                        
	TEL: 297-7197                         <MARIANI.FRED AT A1 AT HOPER AT MRO>
Date:	01-Mar-1991
Posted-date: 01-Mar-1991
Precedence: 1
Subject: AI/NN TOPGUN INVITATION
To:	See Below




                           THE BEST OF THE BEST
_______________   _____   _____   _____   _   _   _   _   _________________
\\\\\\\\\\_   _| |  _  | |  _  | |  _ _| | | | | | \ | | |  ___////////////
  \\\\\\\\\| |   | | | | | |_| | | |     | | | | |  \| | | |___ /////////
    \\\\\\\| |   | | | | |  ___| | |  _  | | | | | \   | |___  |/////// 
      \\\\\| |   | |_| | | |     | |_| | | |_| | | | \ |  ___| |///// 
        \\\|_|   |_____| |_|     |_____| |_____| |_| |_| |_____|///

                      MANUFACTURING CLUSTER MARKETING 
                         FRED MARIANI DTN 297-7197
                                                                                                                               

                           Artificial Intelligence                                   
                                     &
                               Neural Networks 
                                     in 
                         Manufacturing Applications
	                          Workshop


                                  at the

                     Application Center for Technology
                            3945 Freedom Circle
                                 Suite 900
                          Santa Clara, CA 95054

        
                            May 8-10, 1991



                       R.S.V.P. by  April 25, 1991
                      Linda Bersani (508)467 74706



                 TOPGUNS AI & NEURAL NETWORKS IN
                 MANUFACTURING APPLICATIONS WORK
                 
                 Through actual case studies, participants 
                 will learn how others have approached the 
                 integration of tools, technologies, and 
                 people to design, implement, and measure 
                 the positive impact of resultant programs 
                 on manufacturing performance.
                 
                 SAIC, a leading engineering firm in the 
                 field of AI and neural networks, will 
                 present case histories relating to the use 
                 of these technologies in the development 
                 and implementation of intelligent 
                 diagnostics applications.
                 
                 Other case histories will be presented by 
                 MCC (The Microelectronic and Computer 
                 Consortium) and Digital's  own Machine 
                 Learning Consulting Services. These cases 
                 will define the methodologies employed to 
                 use pattern recognition and NN algorithms 
                 to perform the predictive modeling of 
                 complex nonlinear processes, the design of 
                 Six-Sigma process control systems, and the 
                 development of adaptive fault isolation 
                 systems.  Reports of actual levels of 
                 improvement will be provided.
                 
                 From the end-user standpoint, Digital's 
                 Semiconductor Interconnect and Technology 
                 End-User Computing Group will present a 
                 full-day interactive case study detailing 
                 how business pressures driven by the needs 
                 of providing low-cost LSI to the highly 
                 competitive workstation market, forced the 
                 development of AI-based systems, such as 
                 composite wafer mapping, to increase FAB 
                 yields, improve throughput, and reduce 
                 cycle time.  Participants will address 
                 various aspects of the problem including 
                 tool selection, end-user requirements, 
                 cultural changes in the organization, 
                 methodologies, etc..  They will compare 
                 the results of their study to the actual 
                 decisions made and results achieved within 
                 DEC.



                 TOPGUNS AI & NEURAL NETWORKS IN 
                 MANUFACTURING APPLICATIONS WORKSHOP
                 
                                  	    SANTA CLARA ACT
                                             MAY 8-10, 1991
                 
                 
                 
                 AGENDA
                 
                               Wednesday, May 8
                 
                 
                 8:00 - 9:00	Registration and Continental                                                                
                 Breakfast - Santa Clara ACT
                 
                 9:00 - 9:15	Program Introduction
                                Mitch Federman, 
                                Semiconductor Industry Mktg
                 		Digital Equipment Corp
                 
                 9:15 - 5:00 	Case: Digital's Semiconductor                                                 
                                and Interconnect Technology                                                 
                                Group (SCIT Manufacturing)
                 		Bob Moline, Manager SCIT-
                 		End User Computing
                 
                 		Yield Enhancement and                                                 
                                Composite Wafer Mapping 
                 		Venkat Kalkunte, System 
                                Engineer, SCIT-End User
                 		Computing
                 		
                               Thursday, May 9
                 
                 8:00 - 8:30	Continental Breakfast
                 
                 8:30 - 8:45	Program Introduction/Update
                 
                 8:45 - 9:45	Introduction to Neural                                                 
                                Networks and its applications
                                in the semiconductor industry
                 		Phil Wasserman, SAIC
                 
                 9:45 - 10:30	Intelligent Diagnostics                                                 
                                Systems
                 		Timothy Kraft, SAIC
                 
                 10:30 - 10:45	Break
                 
                 10:45 - 12:00	Adaptive Fault Isolation                                                 
                                Systems
                 		Ruby Li, MLCS
                 
                 12:00 - 1:00	Buffet Lunch



                 TOPGUNS AI & NEURAL NETWORKS IN 
                 MANUFACTURING APPLICATIONS WORKSHOP
                 
                 	               	SANTA CLARA ACT
                                         MAY 8-10, 1991
                 
                 
                 
                 AGENDA
                 
                          Thursday May 9 (continued)
                 
                 
                 1:00 - 2:30 	Neural Nets for Supervisory                                                 
                                Process Optimization 
                 		Jim Keeler, MCC
                 
                 2:30 - 2:45	Break
                 
                 2:45 - 4:00	Machine Learning and Decision                                                 
                                Support System in Process                                                 
                                Control
                 		Yuchun Lee, DEC-MLCS
                 
                 4:00 - 4:30	Demo: Yield Enhancement using                                                 
                                Neural Networks
                 
                 4:30 - 5:00	Wrap-up
                 
                 6:00 - 7:30    Dinner (TBD Doubletree?)
                 
                 7:30 -         PANEL DISCUSSION at Hotel
                 
                 
                 
                                Friday, May 10
                 
                 This day has been set aside for one-on-one 
                 consulting with company representatives or 
                 teams who wish advice on the application of 
                 AI and neural nets to their specific 
                 manufacturing problems.  This is a private, 
                 no-cost session.  Sign-up sheets will be 
                 available throughout the workshop, and 
                 individual appointments will be arranged for 
                 convenient times throughout this day.  Take 
                 advantage of this unique service, and 
                 challenge our experts.
                 
                 



                 TOPGUNS AI & NEURAL NETWORKS IN   
                 MANUFACTURING APPLICATIONS WORKSHOP
                 
                 Biographies of Consultants
                 
                 
                 Matti Jaffe, Manager,
                 Machine Learning Consulting Services, 
                 Digital Equipment Corp, Stow, MA
                 
                 Philip Wasserman, Director of Neural Network 
                 Applications,
                 Science Applications International 
                 Corporation, Santa Clara, CA
                 
                 Timothy Kraft, Senior Scientist,
                 Intelligent Systems Division, Science 
                 Applications International Corporation, San 
                 Diego, CA
                 
                 Bob Moline, Manager of End-User Computing
                 Semiconductor Interconnect and Technology 
                 Group,
                 Digital Equipment Corporation, Hudson, MA
                 
                 Venkat Kalkunte, System Engineer
                 Semiconductor Interconnect and Technology 
                 Group,
                 Digital Equipment Corporation, Hudson, MA
                 
                 Ruby Li, Software Engineer and Consultant, 
                 Machine Learning Consulting Services, 
                 Digital Equipment Corporation, Stow, MA
                 
                 Yuchun Lee, Software Engineer and 
                 Consultant, Machine Learning Consulting 
                 Services, Digital Equipment Corporation, 
                 Stow, MA
                 
                 Jim Keeler, Senior Member of Technical 
                 Staff, Advance Computer Architecture and 
                 Systems, Microelectronics and Computer 
                 Corporation, Austin, TX
                 
                 		
                 
                 



                 AUDIENCE
                 
                 
                 *  CIM Managers, Systems Managers, or 
                    Project Managers considering the use
                    of A/I and Neural Networking tools in
                    process control, quality management,
                    or diagnostic applications
                 
                 
                 
                 
                 
                 
                 
                 
                 1991 TOPGUNS
                 PROGRAM SCHEDULE
                 
                 
                 Rapid Prototype
                 June 5-7, 1991
                 Maynard, MA
                 
                 Distribution Requirements
                 Planning/EDI
                 July, 1991 
                 Boston, MA
                 
                 JIT/TQC
                 August 21-23, 1991
                 Colorado Springs, CO
                 
                 CALS/C.E.
                 September 4-6, 1991
                 Irvine, CA
                 
                 Concurrent/Quality Engineering
                 November, 1991
                 Seattle, WA
                 
                 Achieving Total Quality Mgmt
                 December, 1991
                 Dallas, TX
                 
                 
                 
                 






                 HOW TO REGISTER
                 
                 
                 Please confirm your attendance at the TOPGUNS 
                 A/I Neural Networks in Manufacturing 
                 Applications workshop by calling Ms. Linda 
                 Bersani (508) 467-7470 no later than April 24, 
                 1991.
                
                


                

To Distribution List: <<deleted thousands of names>>

Distribution:

TO:  Pat Roach@VBE
TO:  Susan Sugar@MWO
TO:  Steve Becker@AQO
TO:  Ed Hurry@DVO
TO:  SHIRLEY CRIDER@DVO
TO:  STEVE DONOVAN@DLO
TO:  DENNIS DICKERSON@DLO
TO:  Gale Kleinberger@HSO
TO:  Mike Sievers@HSO
TO:  Mike Willis@HSO
TO:  Sherry Williams@HSO
TO:  Katherine Jones@HSO
TO:  Dale Stout@HSO
TO:  Tommy Gaut@HSO
TO:  Tom Wilson@HST
TO:  jim rather@HSO


67.130Interesting Conference this Thurs and Fri - you'reULYSSE::ROACHTANSTAAFL !Wed Apr 10 1991 12:56137
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     10-Apr-1991 10:11am CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Interesting Conference this Thurs and Fri - you're invited!

The Advanced Services Pilots and Prototypes group (ASPPG) invites you to 
attend a sequence of presentation and discussions centering on the use 
of radio frequency networks, LAPTOP computers and knowledge management 
technology. This is a working session to share plans, concepts, goals and 
ideas. We have arranged for some key vendors to present their products and 
answer questions, however the audience is limited to Digital personnel. 
The session is very informal and meant to share concepts only. All material 
is deemed "work in process" and will not be widely disseminated within 
Digital. The conference is open to any DEC employee. 

THEME OF THE CONFERENCE 

When otherwise distinct sequences of technological evolution converge 
in a manner that opens whole new vistas, fundamental changes are required 
in the way we accomplish our objectives. Those who hold onto outdated 
methods and ideas will be left behind. We in ASPPG believe that the 
trends in LAPTOPS, Radio Frequency Networks and Knowledge/Information 
Management Technology pose such an opportunity ... and risk, for Digital. 

Historically the world has experienced two great leaps in productivity: 

           1. The Industrial Revolution

           2. The emergence of free market capitalism

The next leap in productivity will be based on information and knowledge
technology.  

VISION

"A SMALL BOX THAT CAN BE CARRIED IN ONE HAND, THAT IS ABLE TO PROVIDE 
THE KNOWLEDGE NEEDED AT THE TIME AND IN THE FORM NEEDED."

The simplest statement of the requirements to achieve this vision is: 

             1. A very small, but powerful computer that can be 
                carried and used with one hand. 

             2. A means of connecting this unit into a network without 
                constraining the movements of the user.  

             3. Knowledge management software that can capture, store 
                protect and make available the right knowledge at the 
                right time, with minimal requirement for the user to 
                identify the knowledge required. 

The theme of the conference is "MAKING THE RIGHT INFORMATION AVAILABLE, 
AT THE RIGHT TIME, IN THE RIGHT PLACE, IN THE RIGHT FORM, TO THE RIGHT 
PERSON." This is a tall order. Information or knowledge storage implies 
a knowledge representation methodology and database technology. Access 
requires telecommunication networks and presentation software. Security 
is a key component of any such system. The problem of knowledge 
capture needs to be solved. And finally all these elements must be 
present in an integrated infastructure that is easy to redirect and 
use. This last feature may be the most elusive. 




The conference will be held in the Pacific Conference room in DLB5 located 
in Marlboro. 

             LOCATION:      Marlboro, DLB12, Pacific Conference room 

             DATE:          APRIL 11-12, 1991

             START:         APRIL 11, at  9:00AM 

             END:           APRIL 12, at  5:00PM

       AGENDA: 

       Thursday APRIL 11

       9:00-9:05       TERRY SMITH ..........INTRODUCTION 

       9:05-9:30       GARY PERWAK ..........ROLE OF ASPPG 

       9:30-10:15      RICHARD LANG .........INFORMATION UTILITY 

       10:15-10:30     Break 

       10:30-11:15     TERRY SMITH ..........RF/LAPTOPS as SERVICE TOOL 

       11:15-12:00     JACK NELSON ..........RF/LAPTOPS and 
                                                     INTEGRATION BUSINESS

       1:00-1:45       JOE SCHATZ ...........MULTI-MEDIA ON LAPTOPS 

       1:45-2:15       ANDREW FAULKNER ......HUSKY INC. 

       2:15-3:00       MIKE SOHA ............RF NETWORK INVESTIGATIONS 

       3:00-3:15       Break 

       3:15-4:00       Steve Hoyt ...........Knowledge Management Program 

       4:00-5:00       Panel Discussion (Information Utility & Digital)

       Friday, APRIL 12


       9:00-10:15      JEFF MILLER ..........INFORMATION NETWORKING

       10:15-11:00     DEAN SAMMONS .........ARDIS INC. 

       11:00-11:15     BREAK 

       11:15-12:00     KEN KLOSINKI .........MOTORLOA INC. 

        1:00-1:45      Jack Rahaim ..........Knowledge Asset Management

        1:45-2:30      Chuck Marshall .......IMKA Technology

       2:30-3:15       GEORGE FRAZIER .......APPLE COMPUTER INC. 

       3:15-3:30       BREAK 
 
       3:30-4:15       Philip Bourgeault ....RAM Inc. 

       4:15-5:00       Panel Discussion (Potential of RF at Digital)   

67.131YESULYSSE::ROACHTANSTAAFL !Thu Apr 11 1991 11:0845
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     10-Apr-1991 08:10pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: BROWN BAG THIS FRIDAY - AI IN MFG AUTOMATION ARCHITECTURE

			Brown Bag Seminar on
	
		 AI in Manufacturing: Automation Architecture
	

	PRESENTER:	Mr.Atanu Mukherjee
			EIS Manager for Process Industries
			Digital Equipment (India) Ltd
	
	DATE:		Friday, 12 April, 1991

	TIME:		11:30 AM - 1:00 PM
	
	PLACE:		Discovery Conference Room, DLB12-2/D1    
	
	HOST: 		Ram Josyula, SELECT/Fellowship

	Outline: Atanu will be giving a brief presentation on an
	Automation Architecture which deals with planning, scheduling,
	control and integration of manufacturing systems and the role of 
	AI in solving these issues.  He will share his experience in metals 
	industry to illustrate 	his points.

	Atanu is currently visiting Digital's Metals Industry Resource
	Center in Pittsburgh and AITC in Marlboro. He is responsible for
	starting a Digital Competency Center for Metals Industry in India.	
	
	Feel free to forward this memo.
	

67.132Intro to C for people with little-to-no programmingULYSSE::ROACHTANSTAAFL !Fri Apr 12 1991 11:5882
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Apr-1991 10:03pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Intro to C for people with little-to-no programming experience - this      announcement just received

                    INTRO TO STRUCTURED PROGRAMMING IN "C"

                                PART 1

         COURSE NUMBER:     L913302
         DATE:              APRIL 22-26, 1991
         TIME:              8:30 - 5:00
         LOCATION:          T&N TRAINING ROOM TAY2-1
         INSTRUCTOR:        Dr.HARVEY DEITEL/INFORMATIVE STAGES
         COST:              $1000
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
  WHO SHOULD ATTEND:
  This course is specifically for employees with LITTLE OR NO PRIOR
  PROGRAMMING EXPERIENCE who have the need or desire to learn "C".
  It is a friendly introduction to structured programming for novices 
  with emphasis on control structures and arrays.  Approach uses
  examples of complete small programs.

  * [Employees with previous programming experience who need to learn
    "C" should enroll in "Introduction to C for Experienced Programmers"

  COURSE DESCRIPTION:
  This course parallels the first course in computer programming
  taught at the university level. It covers the fundamentals of computer
  programming in the context of the 'C' programming language. It is unique
  in that 'C' has not been taught previously as a first programming language.
  
  FORMAT:
  It is a lecture format with in-class hands-on labs.
  
  COURSE OUTLINE
    o  Lexical elements, operators, and the "C" System
    o  Fundamental Data Types
    o  Flow of Control
    o  Functions
    o  Arrays

  COURSE OBJECTIVES: 
   The attendee will develop a fundamental understanding of computer
   programming concepts, and will be able to write simple programs in C. The
   attendee will be prepared to read and comprehend more complex C prgrams.
                  
INSTRUCTOR:
Harvey Deitel has 30 years experience in the computer field.
He participated in the research and development of several large-
scale operating systems and in the design and implementation of
numerous commercial systems.  He received the Bachelor of Science and
Master of Science Degrees from the Massachusetts Institute of Technology
where he did extensive development work on the Multics operating system.
He received the Doctor of Philosophy Degree from Boston University where
his disseration research examined the problems of developing very large
scale, structured software systems.

Dr. Deitel was previously the chairman of the Computer Science Faculty 
of Boston College where he taught courses in systems programming, 
computer architecture, operating systems, software engineering, and computer 
networking.  He has received numerous teaching commendations, and has been
rated nationally among the top computing teachers in the country. 
He is currently an independent consultant specializing in open systems. 
He is the author of a dozen books including an operating systems textbook 
used in 1000 universities worldwide. Harvey teaches regularly for 
Telecommunications & Networks and receives outstanding evaluations.
 

67.133Advanced C: Algorithms + Data StructuresULYSSE::ROACHTANSTAAFL !Fri Apr 12 1991 12:00127
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Apr-1991 10:47pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Advanced C: Algorithms + Data Structures

                         ADVANCED C WORKSHOP: ALGORITHMS & DATA STRUCTURES

         COURSE NUMBER:  L915601
         DATE:           MAY 7-10, 1991
         TIME:           8:30 - 5:00
         LOCATION:       T&N TRAINING ROOM TAY2-1
         INSTRUCTOR:     DAVID PROSSER/TECHNOLOGY EXCHANGE COMPANY
         COST:           $1200
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------
Prerequisite:
Must have two years of professional programming in C.

Intensive four days of programming.

Course Overview
This course provides attendees with a solid foundation in data structures
and algorithms enhancing an individual's development of portable,
efficient and maintainable C code.  Course work and exercises are developed
to build progressively upon programming tasks culminating in a large and 
useful program. Throughout the course, a "Toolbox" approach to the activities
will be emphasized, augmenting the student's personal programming repertoire.  
This course is intended for application and systems programmers who know C,
but participants need not be experts in it or ANSI C. 

Attendee Qualifying Questions:
1.  Have you been actively programming in C for at least 2 years?
2.  Are your programs at least 500-1000 lines long?
3.  Do you regularly use pointers?
4.  Are you comfortable with dynamic allocation/deallocation?
5.  Do you consider writing recursion?
6.  Do you break your program into modules and header files?
7.  Do you think about portability and efficiency of your code?

If you answer yes to these questions this course is appropriate for you.
If not you will find the course difficult.  This questionnaire will be
sent to enrollees.

Course Content
Introduction
Overview
   o  Assumed prerequisites and skills
   o  "Toolbox" approach to programming
   o  All programming in C
Goals Gain:
   o  Advanced C programming skills
   o  Experience with basic and advanced algorithms and data structures
   o  Skills in practical and balanced programming
   o  An enhanced programming repertoire
Review of C
ANSI C
   o  Quick history of C and ANSI C
   o  How to write programs that work with both K&R and ANSI C
   o  How to compile, link, and run programs
Types
   o  Basic type  o  Derived types  o  Incomplete types
Expressions
   o  Precedence  o  Constant expressions  o  Initialization
Preprocessing
   o  Header inclusion  o  Macros  o  Conditional inclusion
Libraries
   o  Input/output  o  String operations
   o  Character handling  o  Reserved name space
Object-oriented programming
   o  Information hiding  o  Abstract data types  o  C++
Advanced C Programming
Balanced programming 
   o  Portability  o  Maintainability  o  Efficiency  o  Reusability
Implementation choices
   o  recursion/iteration  o  Dynamic/static
   o  Memory/file storage  o  General/specific
Cost analysis
   o  Big "O"  o  Time/space  o Measurement
Algorithms and Data Structures
Algorithms
   o  Searching  o  Sorting  o  Inserting  o  Deleting  o  Hashing
Data structures 
   o  Lists  o  Sets  o  Trees  o  Graphs
Lists
Data types
   o  Stacks  o Queues  o  Strings  o  Circulars  o  Mappings
Operations
   o  Push/pop  o  Searching  o  Sorting  o  Insertion  o  Deletion
Sets
Data types
   o  Simple  o  Tables  o  Bags
Operations
   o  Union  o  Intersection  o  Difference  o  Find  o  Insert  o  Delete
Trees
Data types
   o  Binary  o  Balanced  o  N-ary
Operations
   o  Traverse  o  Sort  o  Insert  o  Delete  o  Construct  o  Parse
Graphs
Data types
   o  Directed  o  Acyclic  o  Undirected
Operations
   o  Traverse  o  Construct
Programming Techniques
Divide and conquer
Backtracking
Dynamic programming

Instructor:  David f. Prosser
Dave Prosser is an expert on the C programming language.  He is the 
editor of the new ANSI C standard as well as serving as AT&T Bell Laboratories's
representative to the standardization committee.  He has been programming in 
C since the late 1970s and has been part of Bell Lab's C language development 
area for almost ten years.

67.134Satellite Course Calendar - Apr-NovULYSSE::ROACHTANSTAAFL !Fri Apr 12 1991 12:05109
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     11-Apr-1991 10:47pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Satellite Course Calendar - Apr-Nov

From the NTU Satellite Network:  The following satellite courses are 
			 	 currently scheduled at LKG

  ------------------------------------------------------------------------
	  For course descriptions and to register:
	     
	  Set Host DELNI; Username: COURSES; Password: not required
          Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
  ------------------------------------------------------------------------

Apr.16-17  Digital Video Transmission:     
             Algoritms to End-Products
12 Hrs.    Instructor: Dr. Winthrop W. Smith and Others
S915001    NTU Satellite Network: Live Satellite Broadcast

Apr.19     Linking Pcs in the 90s Using Novell
6 Hrs.     Instructor: Paul J. Dattoli & James P. Stewart
S915501    NTU Satellite Network: Live Satellite Broadcast

Apr.26,    Data Communications II  
May 3,10 
18 Hrs.    Instructor: Eugene von Taube
S914501    Network Northeastern: Live Satellite Broadcast

Apr.26     Object-Orineted Design Using C++   
6 Hrs.     Instructor: Dr. Ira Pohl
S915201    NTU Satellite Network: Live Satellite Broadcast

Apr.30     An Introduction to Software Metrics  
6 Hrs.     Instructor: Denis Meredith
S915401    NTU Satellite Network: Live Satellite Broadcast

May 14-15  Data Communications Network Decision-Making  
12 Hrs.    Instructor: Dr. H. Charles Baker
S915101    NTU Satellite Network: Live Satellite Broadcast

May 17     Theory of Object-Oriented Design and its 
             Application to C++                    
6 Hrs.     Instructor: Dr. karl Lieberherr
S915301	   NTU Satellite Network: Live Satellite Broadcast

May 20     Telecommunications Technolgies Update:  
             (BISDN) Broadband Integrated Services Digital Network 
3 Hrs.     Instructor: Bob Bovenizer and Terry Curtis
S916001    NTU Satellite Network: Live Satellite Broadcast

May 23     IEEE Video Conference: Neural Network Applications 
             for the 1990s                         
3 Hrs.     Instructor: James A. Anderson, Dan Hammerstrom, Larry Jackel
S915601    NTU Satellite Network: Live Satellite Broadcast

June 11    C Programming For Engineers: Introduction   
6 Hrs.     Instructor: Dr. F. S. Hill, Jr.
S915701    NTU Satellite Network: Live Satellite Broadcast

June 14,   Data Communications III     
21,28
18 Hrs.    Instructor: Eugene von Taube
S914601    Network Northeastern: Live Satellite Broadcast

June 17    Telecommunications Technologies Update:
             Class Services                         
3 Hrs.     Instructor: Dennis Squires
S916201    NTU Satellite Network: Live Satellite Broadcast

June 25    C Programming for Engineers: Advanced  
6 Hrs.     Instructor: Dr. F.S. Hill, Jr.
S915801    NTU Satellite Network: Live Satellite Broadcast

June 26    Interoperability: The Concepts and Technologies
             of Open Systems Networking             
6 Hrs.     Instructor: Ronald G. Corriveau
S915901    NTU Satellite Network: Live Satellite Broadcast

July 15    Telecommunications Technologies Update:
             SMDS: A Broadband Service                 
3 Hrs.     Instructor: Donald Dineen
S926101	   NTU Satellite Network: Live Satellite Broadcast

Aug.8,     Network Management  
15,22
18 Hrs.    Instructor: Scott Helmers
S914701    Network Northeastern: Live Satellite Broadcast

Oct.4,11,  Fundamentals of OSI and DOD Protocols  
18
18 Hrs.    Instructor: Ron Corriveau
S914801    Network Northeastern: Live Satellite Broadcast

Nov.8,22,  Voice Data Planning Seminar
Dec.6
18 Hrs.    Instructor: Dr. James A. Metzler
S914901    Network Northeastern: Live Satellite Broadcast

67.135REMINDER - TSS re. DICE 4/25, 10AMULYSSE::ROACHTANSTAAFL !Wed Apr 24 1991 19:3099
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     24-Apr-1991 06:10pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: REMINDER - TSS re. DICE 4/25, 10AM







	TITLE: 		"Computer-Aided Cooperative Product Development"

	SPEAKER:	DUVVURU SRIRAM
			Assistant Professor and Co-Technical Director
                     	Intelligent Engineering Systems Lab.
                        Dept. of Civil Engineering
                        M. I. T., Cambridge, MA 02139
 
	DATE:		Thursday, 25 April, 1991

	TIME:		10:00 AM - Noon

	PLACE:		Pacific Room, DLB12

	HOST:		GARRY WOODRUFF
			DASH Program Manager
			ISTG/Concurrent Engineering Group

	Computer aided tools, which will be  collectively  called  DICE
	(Distributed  and  Integrated  environment for Computer-aided 
	Engineering), are currently being developed using concepts from 
	databases and AI.  

	In this  talk, various  features of the DICE framework will be
	described, such as an object-oriented blackboard, transaction and 
	version management modules and a shell for preliminary design.

	This is being done to address the following problems:  
		
		o Typical engineering projects involve a large number 
	          of components and the interaction of multiple tech-
		  nologies.  

		o On a single project, interacting design technologies 
	          often come from separate firms or functional groups 
	          within a firm, and there is little coordination 
		  between various participants.  

		o Typically, engineers find coordination among themselves 
	          difficult and they leave this task to the manufacturing 
		  personnel.  

		o Thus, working drawings,  used to inform the fabricator 
	          of the product, lack detail.   Hence, potential conflicts 
		  among participants are often unrecognized until manufac-
	          turing begins.
    		
		o Several undesirable effects are caused by this lack of 
	          coordination.







	Duvvuru Sriram TSS
	Page Two

	   

	Duvvuru Sriram is an assistant professor of Civil Engineering and  
	the co-technical director of the Intelligent Engineering Systems  
	Laboratory  at MIT.  

	Currently, he is working in the areas of knowledge-based systems,
	development of databases for engineering design, and natural 
	language processing.  

	He was a founding co-editor of the International Journal for AI 
	in Engineering.  He has nearly 70 publications to his credit, which  
	includes  two books and eight edited volumes on AI in engineering.  
	Sriram is also a member	of the advisory board for several inter-
	national conferences and journals.   In 1989, he was awarded a 
	Presidential Young Investigators award from the National Science 
	Foundation.  Sriram received a B. Tech (1980) degree  from  I.I.T.,
	Madras, India, and MS (1981) and Phd (1986) degrees from CMU.

67.136CRL Seminar, 5/15, Ken NakayamaULYSSE::ROACHTANSTAAFL !Thu Apr 25 1991 11:3686
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     24-Apr-1991 06:25pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: CRL Seminar, 5/15, Ken Nakayama

From:	AIADM::CRL::MARYG "Mary Gallagher" 24-APR-1991 11:32:03.85
To:	aiadm::tss 
CC:	
Subj:	Visual Surface Perception - Digital Cambridge Research Lab - May 15 

                     Digital Equipment Corporation
                        Cambridge Research Lab 
                               Seminar
 
                         Wednesday, May 15, 1991
                  3:15pm - 4:15pm, Refreshments at 3:00pm
 
	               "Visual Surface Perception"
 
			     Ken Nakayama
		       Department of Psychology
			  Harvard University
 
 
Because we live in a 3 dimensional world, along any given general
direction, our visual system is often confronted with multiple surfaces,
not just a single surface.  We make the working assumption that our
visual system and those of other animals have developed (possibly through
learning), very specific and probably "low level" mechanisms for dealing
with these ever-present and unavoidable situations.  Thus we think that
occlusive relations in the world are processed very early in the visual
system of primates, probably as early as the primary visual cortex.  In
addition, we propose the "generic view principle" to understand why the
visual system encodes: 1) surfaces continuing behind other surfaces, 2).
surfaces completing accross other surfaces (as in subjective contours),
3) perceived transparency in very simple stereograms.
 
A set of visually unexpected and surprising stereoscopic demonstrations 
will accompany the talk.
 
-------------------------------------------------------------------------------
 
Ken Nakayama is Professor of Psychology, Faculty of Arts and Sciences at
Harvard University.  He received his Ph.D. degree from U.C.L.A in the
late 1960s.  After doing a post-doctoral fellowship at U.C. Berkeley, and
teaching at a medical school in Canada, he spent nearly the last 20 years
as a senior scientist at Smith Kettlewell Eye Research Institute in San
Francisco.  His early interests and work concentrated on low-level
vision, including the processing of neural signals in the retina and
lateral geniculate as well as studying eye rotations.  More recently, he
has become interested in intermediate level vision, using psychophysical
methods of investigation.
 
-------------------------------------------------------------------------------
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and 
Broadway.  The address is:
 
     Digital Equipment Corporation
     Cambridge Research Lab
     One Kendall Square, Bldg. 700
     Cambridge, MA 02139
 
For questions regarding the seminar, directions, or parking, or to add 
your name to the mailing list, call DTN: 259-6601 or send email to 
crl::tss
 
 
% ====== Internet headers and postmarks (see DECWRL::GATEWAY.DOC) ======
Received: by easynet.crl.dec.com; id AA08904; Wed, 24 Apr 91 08:49:22 -0400
Message-Id: <[email protected]>
To: aiadm::tss
Subject: Visual Surface Perception - Digital Cambridge Research Lab - May 15
Date: Wed, 24 Apr 91 08:49:28 EDT
From: maryg

67.137FWD: IMKA USER'S FORUM - MAY 13, 1991ULYSSE::ROACHTANSTAAFL !Mon May 06 1991 11:1149
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     03-May-1991 05:15pm CET
                                        From:     BEANE
                                                  BEANE@BIGRED@MRGATE@DPD04@DPD
                                        Dept:      
                                        Tel No:    

TO: See Below

Subject: FWD: IMKA USER'S FORUM - MAY 13, 1991

From:	GUESS::SANDOCK "Carmel @ LMO2 1/L3 - 296-5102  02-May-1991 1547"    2-MAY-1991 15:01:21.88
To:	@IMKA_INTEREST.DIS
CC:	SANDOCK
Subj:	IMKA USER'S FORUM - MAY 13, 1991

Attached you will find the agenda for the IMKA User's forum which is 
taking place on May 13.  

Please respond to GUESS::SANDOCK by May 6 if you plan to attend or have
any questions.


			IMKA USER'S FORUM

			  May 13, 1991
			  2:00 - 3:30
		     Leopard c/r  DLB5 3/E10

		
		2:00 - 2:15 INTRODUCTION       Dino Lachiusa

		2:15 - 2:45 IMKA Technology    Chuck Marshall

                2:45 -       Open Discussion         


Distribution:


TO:
DENNIS DICKERSON@DL STEVE DONOVAN@DLO   Tommy Gaut@HSO      Ed Hurry@DVO        
Louis Pau@VBE       jim rather@HSO      Pat Roach@VBE       Czarena Siebert@HSO 
Mike Sievers@HSO    Dale Stout@HSO      Susan Sugar@MWO     Sherry Williams@HSO 
Mike Willis@HSO     Tom Wilson@HST      

67.138Neural Nets - 1 Hour Video TutorialULYSSE::ROACHTANSTAAFL !Sun May 12 1991 16:09100
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-May-1991 06:12pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: Neural Nets - 1 Hour Video Tutorial

--------------------------------------------------------------------------------
                  ETE Satellite Network Courses Description
________________________________________________________________________________
Title             Neural Network Applications for the 1990s                   
Instructor        Various speakers              
Date              23-MAY-91 - 23-MAY-91
Time              11:00 -  3:00
Location          HLO, APO, FXO                 
Course Number     14SAT-61
Course Fee        $  300
Registration      Log into COURSES software on SHARE  
                  Username is COURSES and password is EDUCATION.
Cancellation      Please withdraw 1 week  prior to start date to avoid being  
                  charged                                                     


                          *** SPECIAL ONE HOUR VIDEO TUTORIAL ***
                                 11:00 AM - 12:00 PM ONLY
                      *** PROGRAM WILL RUN FROM 12:00 PM - 3:00 PM ***
                                                                                
Note              Room locations: HLO- UVC Classroom HLO2-1/D08
                  APO-V135   FXO-To be determined
                                                                                
Description       This videoconference will focus on near-term and far-term
                  applications for neural network technology.  Although
                  applications have now started to appear, this novel
                  technology will not start to deliver its considerable
                  promise until later in the decade.  Therefore this
                  videoconference will necessarily be somewhat speculative, in
                  addition to presenting current accomplishments and plans for
                  the immediate future.
                                                                                
Outline           Introduction (5 minutes)
                  JAMES ANDERSON
                  *  Rudiments of the massively parallel neural network
                     technology
                  *  Terminology of the field
                                                                                
                  DAN HAMMERSTROM
                  Present and Future Plans to Build Digital VLSI Circuitry (40
                  minutes)
                  *  Design considerations peculiar to neural net VLSI
                     architectures
                  *  Examples of specific designs and why they were chosen
                                                                                
                  LARRY JACKEL
                  Current Applications of Neural Networks (40 minutes)
                  *  Current neural network applications, including AT&T Bell
                     Laboratories' character recognition system (neural network
                     recognition of hand-printed zip codes).
                  *  Speech recognition, signal processing, financial data
                     processing and optimization.
                                                                                
                  JAMES ANDERSON
                  Current and Future Applications (40 minutes)
                  *  Need for large networks
                  *  Need for architectures for networks used in certain
"cognitive"             
                     operations.
                                                                                
                  Examples such as radar classification system, network
                  model for object motion, and network capable of
                  performing some operations in elementary arithmetic.
                                                                                
                  *  Speculations about networks
                                                                                
                     Artificial systems with "intuition" and "common sense"
                     Hybrid systems, part neural network and part traditional
                     artificial intelligence
                  *  Computational limitations of current neural network


                     models
                  *  Important criticisms of Fodor and Pylyshn about
                     computational adequacy of current network formulations.
                                                                                
Speakers          James A. Anderson is a professor in the Department of
                  Cognitive and Linguistic Sciences and Department of
                  Psychology at Brown University. Dan Hammerstrom is chief
                  technical officer of Adaptive Solutions, Inc., of Beaverton,
                  Ore., and an associate professor at the Oregon Graduate
                  Institute.  Larry Jackel is with AT&T Bell Laboratories in
                  Holmdel, N.J.
 

67.139ULTRIX Fundamentals Course OfferingULYSSE::ROACHTANSTAAFL !Sun May 12 1991 16:11158
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-May-1991 06:12pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: ULTRIX Fundamentals Course Offering

       ULTRIX FUNDAMENTALS: AN INTENSIVE INTRODUCTION TO UTILITIES &
                            COMMANDS AND SHELL PROGRAMMING


         COURSE NUMBER:    L912908
         DATE:             MAY 28-31, 1991
         TIME:             8:30 - 5:00
         LOCATION:         T&N TRAINING ROOM TAY2-1

         COURSE NUMBER:    L912909
         DATE:             JUNE 25-28, 1991
         TIME:             8:30 - 5:00
         LOCATION:         FREEPORT CLASSROOM LKG2-2/V4
         INSTRUCTOR:       DAVE RUBIO/RUBIO ASSOCIATES
         COST:             $800    
         
------------------------------------------------------------------------
To Register:  Set Host DELNI; Username: COURSES; Password: not required
              Or Contact Faye Napert; DTN: 226-7065; DELNI::ED_SERVICES
------------------------------------------------------------------------

ULTRIX FUNDAMENTALS: AN INTENSIVE INTRODUCTION TO UTILITIES & COMMANDS
& SHELL PROGRAMMING

COURSE GOALS:
This course is desgined for the new Ultrix user.  Its goal is to
familiarize the student with the general capabilities of the Ultrix
operating system.  The student will learn to create and modify files
using the text editor vi; manipulate files and directories using
shell (csh) comands; become familiar with file permissions; execute
simple, sequential, and pipeline command lines; and communicate with 
other users.  The student will become familiar with the standard Ultrix
documentation.

KEY TOPICS:
Passwords
File system
File permissions
chmod
Ultrix shell
vi text editor
Regular expressions
Shell meta-characters
I/O redirection
Manual pages 
Pipelines
mail

FORMAT:
Lecture plus Lab exercises

MATERIALS:
Student workbook
Textbook "Introducing the Unix System"

RREQUISITES:
Familiarity with computer system concepts and terminals.


                 COURSE OUTLINE - ULTRIX FUNDAMENTALS

I. INTRODUCTION
   Ultrix features
   Ultrix layering
   Ultrix documentation

II.LOGGIN IN
   Setting your terminal
   Password command
   Ultrix shell
   Ultrix shell
   Ultrix file system

III.COMMUNICATION COMMANDS

   mail
   who
   write
   news

IV.ULTRIX FILE SYSTEM
    File types
    Directories
    Pathnames
    File permissions
    File system commands

V.INTRODUCTION TO VI EDITOR
   Cursor Movement
   Reading/Wtiting Files
   Buffer
   Commands
   Regular expressions
   Searching
   Substituting

VI.ESTABLISHING A LOGIN ENVIRONMENT
   Environment variables
   .cshrc
   .login
   .profile

VII.ULTRIX SHELL
   Features
   Shell meta-charactes
   Command lines

VIII.I/O REDIRECTION AND PIPELINES
   Standard input and output
   Standard error
   Redirection
   Pipelines

IX.FILE SYSTEM COMMANDS
   cp,mc,ln
   ls
   find
   cpio
   tar
   file
   od

X.TEXT PROCESSING COMMANDS
  grep
  sort
  diff
  wc
  pr
  tail
  spell

XI.BACKGROUND COMMAND LINES
  Process ids
  ps
  wait
  kill
  nohup
  Job Control

XII.OVERVIEW OF ULTRIX ENVIRONMENT
  Development tools (compiling C programs)
  Text Processing tools

67.140CRL Seminar - "Visual Surface Perception" 5/15 @3:1ULYSSE::ROACHTANSTAAFL !Sun May 12 1991 16:1271
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     08-May-1991 07:13pm CET
                                        From:     TSS
                                                  TSS@AIADM@SELECT@HERON@ULYSSE@MRGATE@VALMTS@VBO
                                        Dept:      
                                        Tel No:    

TO:  PATRICK ROACH@VBE


Subject: CRL Seminar - "Visual Surface Perception" 5/15 @3:15PM

                     Digital Equipment Corporation
                        Cambridge Research Lab 
                               Seminar
 
                         Wednesday, May 15, 1991
                  3:15pm - 4:15pm, Refreshments at 3:00pm
 
	               "Visual Surface Perception"
 
			     Ken Nakayama
		       Department of Psychology
			  Harvard University
 
Because we live in a 3 dimensional world, along any given general
direction, our visual system is often confronted with multiple surfaces,
not just a single surface.  We make the working assumption that our
visual system and those of other animals have developed (possibly through
learning), very specific and probably "low level" mechanisms for dealing
with these ever-present and unavoidable situations.  Thus we think that
occlusive relations in the world are processed very early in the visual
system of primates, probably as early as the primary visual cortex.  In
addition, we propose the "generic view principle" to understand why the
visual system encodes: 1) surfaces continuing behind other surfaces, 2).
surfaces completing accross other surfaces (as in subjective contours),
3) perceived transparency in very simple stereograms.
 
A set of visually unexpected and surprising stereoscopic demonstrations 
will accompany the talk.
 
-------------------------------------------------------------------------------
 
Ken Nakayama is Professor of Psychology, Faculty of Arts and Sciences at
Harvard University.  He received his Ph.D. degree from U.C.L.A in the
late 1960s.  After doing a post-doctoral fellowship at U.C. Berkeley, and
teaching at a medical school in Canada, he spent nearly the last 20 years
as a senior scientist at Smith Kettlewell Eye Research Institute in San
Francisco.  His early interests and work concentrated on low-level
vision, including the processing of neural signals in the retina and
lateral geniculate as well as studying eye rotations.  More recently, he
has become interested in intermediate level vision, using psychophysical
methods of investigation.
 
-------------------------------------------------------------------------------
Digital Equipment Corporation's Cambridge Research Lab is located on the
north side of Technology Square, at the junction of Hampshire Street and 
Broadway.  The address is:
 
     Digital Equipment Corporation
     Cambridge Research Lab
     One Kendall Square, Bldg. 700
     Cambridge, MA 02139
 
For questions regarding the seminar, directions, or parking, or to add 
your name to the mailing list, call DTN: 259-6601 or send email to 
crl::tss

67.141Dynamic Business Modeling with Symmod - Course AnnoMR4DEC::ROACHTANSTAAFL !Tue Dec 03 1991 20:41109

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Doc. No:  006757
                                        Date:     03-Dec-1991 02:36pm EST
                                        From:     LEVIN
                                                  LEVIN@LMOADM@MR4DEC@MRGATE@SALES@MRO
                                        Dept:      
                                        Tel No:    

TO:  PAT ROACH@MRO


Subject: Dynamic Business Modeling with Symmod - Course Announcement 

VMSmail To information: @CC$PUBLIC:MASTER.DIS
Sender's personal name: Suegene Levin, AI Training Registrar, DTN 291-8600  03-Dec-1991 1348


                     DYNAMIC BUSINESS MODELING WITH SYMMOD

         DATES:          DECEMBER 9-13, 1991
			 JANUARY 27-31, 1992
			 
         TIME:           9:00 - 5:00

         LOCATION:       MEDITERRANEAN ROOM, DLB12-1


         COST:           $1600

	 CANCELLATION:	All cancellations must be made at least ten working
                        days before the start of class.

------------------------------------------------------------------------
To Register:  Contact Suegene Levin, DTN: 291-8603 or AIADM::LEVIN
------------------------------------------------------------------------



	Dynamic Business Modeling is a way to test business process 
	designs before they are implemented.  Dynamic Business Modeling 
	can help a business uncover opportunities in existing business
	process, design new processes, and develop transition plans to 
	help the business get there.  Dynamic Business Modeling is an 
	essential component of business process re-engineering, 
	developing systems and business integration proposals, 
	or the implementation any large-scale change project.
	

Course Content:

	This 5 day course will cover Dynamic Business Modeling
	with emphasis on the development of business process models 
	using SYMMOD.  The course includes hands-on lab work in developing 
	and presenting models.
    
	The course also includes an overview of Dynamic Business Modeling,
	business process design and its importance in large-scale
	change projects, and knowledge aquisition for modelers.

	This course is designed for anyone whose role in business 
	process design.


Objectives:

	You will learn:

		o where Dynamic Business Modeling is applied

		o Dynamic Business Modeling's role in business process 
		  re-engineering efforts

		o How to use SYMMOD, the modeling tools that support
		  Dynamic Business Modeling

		o An overview of knowledge acquisition

	

Pre-requisites:

	- The ability to use an editor (e.g. EDT)
	- Experience using DECwindows application is helpful


To Register:	

	Contact Suegene Levin, DTN: 291-8603 or AIADM::LEVIN, with the following
        information:

Name:
Badge No.:
Group Name:
Organization:
DTN:
E-Mail address:
Mailstop:
Manager's Name:
Manager's E-mail:
Cost Center:






67.142IMAGING - IMPLEMENTING SCALABLE SOLUTIONS - CourseMR4DEC::ROACHTANSTAAFL !Tue Dec 03 1991 23:14180

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Doc. No:  006760
                                        Date:     03-Dec-1991 02:58pm EST
                                        From:     LEVIN
                                                  LEVIN@LMOADM@MRGATE@MR4DEC@MRO
                                        Dept:      
                                        Tel No:    

TO:  PAT ROACH@MRO


Subject: IMAGING - IMPLEMENTING SCALABLE SOLUTIONS - Course announcem

                            Training Announcement

          TITLE:           IMAGING - IMPLEMENTING SCALABLE SOLUTIONS

	  AUDIENCE:	   Customers, Sales Support, Sales

          DATES:           Dec 11th & 12th 1991 
			   Feb 18th & 19th 1992 

          TIME:            9:00 - 5:00

          PLACE:	   Room 144a (Andaman)
			   LMO           

          COST:            $900.00 

	  COURSE NUMBER:   DKAS-8613E

          High performance image-based solutions are being used throughout
          the industry to solve specific problems.  This course will
          provide you with the foundation and tools to identify potential
          application opportunities, recognize and characterize problems for
          which a large scale high performance imaging solution is
          appropriate and ascertain a customer's critical requirements.

          IMAGING - IMPLEMENTING SCALABLE SOLUTIONS is a non-product-specific
          course aimed at solutions providing for hundreds of seats, handling
          tens of thousands of documents per day.

          You will learn:

          � What is the current Imaging technology and what hardware and
            software components are  included in a large scale high
            performance imaging solution

          � How to characterize imaging problems and how to determine the 
	    size and scope of a solution

          � What is the architectural strategy required for large scale
            high performance application-centered systems

          � The role and impact of workflow in integrated image systems,
	    including process re-engineering for optimized workflow

	  A more detailed outline is attached.

          Who should attend:

	  Customers accompanied by their Digital Sales and/or Support
	  representatives who are considering large scale applications.

          This course is valuable for Sales Support personnel involved
          in selling large scale, high performance imaging solutions.

          Course Perspective:
	  This course offers a description of a high-end system which ex-   
          amines the technology and solution strategy, rather than specific
          products. The business case is the common thread tying this course
	  together. The material complements but does not duplicate the goals
          of other product specific courses available. 
          
          

          TO REGISTER: Call Suegene Levin DTN 291-8600 or send mail to 
          AIADM::LEVIN




	  Course Outline


              IMAGING: INTEGRATING SCALABLE SOLUTIONS

		                
 		Image Solutions Introduction 
                  o  The Imaging Market Outlook			
                  o  Business Problems & Characteristics
                  o  Approach, Assumptions, Philosophy
                  o  Image System Solution Components 

		Demo of an integrated Image Solution
                  o  Application Template 

		Imaging Technology Introduction 
		  o  Standards & Support
                  o  Image representation 
                  o  Compression & decompression
                     - Hardware & software techniques & trade-offs

		Imaging Systems Architecture 
                  o  Image Solutions
                  o  Client-Server
                  o  High performance 
                  o  Scalability
                  o  Components & their relationship in a system

		Scalable Servers
                  o  Standards compliance 
                  o  Storage 
                  o  Database 
                  o  Utilities

		Components -- A Closer Look 
                  o  End-user Applications
                     - Functionality drives the system configuration
                     - Local processing issues & advantages
                     - Focus of integration 
                  o  Input
                     - Scanning & Quality Control 
                     - FAX			  
                     - Video & other
                  o  Image Storage systems & management
                     - Magnetic caching
                     - Optical storage
		     - WORM
		     - Magnetic Optical
		     - Combined
		     - Helical Tape



              IMAGING: INTEGRATING SCALABLE SOLUTIONS



		Components -- A Closer Look (cont.)
                  o  Database Access & Control 
                     - Reading, Adding & Modifying Data
                     - Linking solution functions 
                     - Transaction monitoring & Scalability
                  o  Output Options
                     - Monitor display
                     - Printer
                     - FAX
                     - Other (EDI, etc)

		  o  Workflow Control

		OCR/ICR
                  o  Optical Character Recognition
                  o  Intelligent Character Recognition


		System Integration & Project Management Issues
                  o  Project Planning & Functional Spec
                  o  Implementation guidelines
                  o  SI issues - Coordination & Control
                  o  Expansion & Migration 

		Business Case 
                  o  Business modeling 
                     - Process re-engineering 
                     - Tools 
                  o  Productivity & Payback
                  o  Segmentation 
		  o  Future


		Panel
                  o  4-6 expert panelists 
                  o  Open to questions from participants


67.143Some DLB training offeringsMR4DEC::ROACHTANSTAAFL !Tue Dec 17 1991 16:24274

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Doc. No:  007046
                                        Date:     17-Dec-1991 04:13am EST
                                        From:     ROACH
                                                  ROACH@AIDEV@MRGATE@NRGATE@NRO
                                        Dept:      
                                        Tel No:    

TO:  PAT ROACH@LMO


Subject: New notes for conference EXPERT                             

------------------------AI Courses at DLB12-------------------------------------
206.2   LMOADM::LEVIN         -< Open Interface >-
206.3   LMOADM::LEVIN         -< KBMS TRAINING >-
206.4   LMOADM::LEVIN         -< NEXPERT OBJECT TRAINING >-

		Notes start on next page.


               <<< BROKE::STAR01:[NOTES$LIBRARY]EXPERT.NOTE;1 >>>
                              -< Expert Systems >-
================================================================================
Note 206.2                     AI Courses at DLB12                        2 of 4
LMOADM::LEVIN "Suegene Levin"                        70 lines  12-DEC-1991 16:20
                              -< Open Interface >-
--------------------------------------------------------------------------------



			Training Announcement


TITLE:		GUI DEVELOPMENT WITH OPEN INTERFACE

AUDIENCE:	Software developers, user interface designers

LENGTH:		5 Days

DATES:		February 10th - 14th
		March 23rd - 27th
		May 18th - 22nd

TIME:		9:00 a.m. - 5:00 p.m.

PLACE:		LMO2, Marlboro, Massachusetts 

COST:		$1695.00


To register for this course, please contact Suegene Levin, LMOADM::LEVIN, DTN
291-8600.

OPEN INTERFACE

Open Interface is a graphical user interface (GUI) development tool developed
by Neuron Data, Inc.*  It contains a WYSIWYG graphics editor and an API
(application programming interface),  a library of C functions used to
manipulate the  windows and widgets in your interface.  The editor, Open
Editor, allows you to graphically layout your interface.  It then generates
resource files which  describe the physical layout and appearance of your
interface (e.g. X & Y  coordinates, color, font) and C templates which act as a
shell which you customize using the API and other C functions to describe the
behavior of your interface.

The applications developed using Open Interface are fully portable over many 
platforms regardless of where they are developed.  The currently available 
platforms are VMS, UNIX, MS DOS, OS/2, and Macintosh.  For example, an 
interface can be developed on the VAX under VMS and ported to the PC by just 
recompiling and linking on the PC.  The applications can easily be changed to 
simulate any of the window managers running on these platforms (MOTIF, MS 
Windows, Presentation Manager, Macintosh, and Open Look).  

This course is a 5 day, highly interactive, hands-on workshop.  The course 
will cover:

  o Learning the basic concepts of window programming

  o Creating user interface layouts using Open Editor

  o Customizing the windows and widgets created in Open Editor using the API 
    and C libraries

  o Integrating Open Interface with other applications

  o Porting the application to different platforms


The workshop format is lecture/lab where you will learn concepts and then
apply them.  By the end of the course the student will have created a
complete application utilizing all the widgets available and most of the 
significant components of the API.

Experience with C is preferred.

*Digital is a certified trainer of Neuron Data products.

================================================================================
Note 206.3                     AI Courses at DLB12                        3 of 4
LMOADM::LEVIN "Suegene Levin"                        79 lines  12-DEC-1991 16:22
                               -< KBMS TRAINING >-
--------------------------------------------------------------------------------
                    ***Announcing KBMS Training Course***

Dates:		Oct 14-18, 1991
		Feb. 17-21, 1992
		June 1-5, 1992

Location:	DLB12, Marlborough, Mass.

Cost:		$1695.00

To register, please contact Suegene Levin, DTN 291-8600, or send the following
information to AIADM::LEVIN -

Name:
Badge No.:
Group Name:
Organization:
DTN:
E-Mail address:
Mailstop:
Manager's Name:
Manager's E-mail:
Cost Center:




Course Description

COURSE:            PROGRAMMING IN KBMS (R)

CURRICULUM:        Knowledge Engineering, Fastrak, Emerging Technologies

DESCRIPTION:       KBMS (R) (Knowledge Based Management System) is an expert 
		   system tool developed by AICorp. KBMS offers high-level,
                   seamless integration with different database
                   management systems and the ability to be
                   called from existing conventional
                   application programs. The KBMS architecture
                   makes it highly flexible for use in
                   developing a wide variety of applications
                   and for enhancing existing product
                   applications. The course covers the use and
                   features of KBMS knowledge base development and use.

GOALS:             The course consists of lectures and labs. During the lectures
                   concepts and features of KBMS are covered.
                   Labs will provide the students with a
                   hands-on experience working with KBMS on VAX
                   workstations. The goals of the course are to
                   enable the participants to:

	           � develop an expert system using KBMS
                     components and inferencing strategies.

	           � retrieve database information using the
                     natural language facilities in KBMS.

	           � effectively use KBMS features for
                     enhancing fluency and rule definition.

	           � efficiently access internal and external
                     object occurrences.

	           � evaluate expert system performance and
                     tune KBMS applications.
                     
PREREQUISITES:     It is assumed that students will come into the course with
                   knowledge of at least one programming
                   language. Previous experience with expert
                   systems is not necessary but may be helpful.

LENGTH:            5 days

FORMAT:            Lecture/Lab

                                                     Revised June 10, 1991

KBMS (R) is a registered trademark of AICorp.
================================================================================
Note 206.4                     AI Courses at DLB12                        4 of 4
LMOADM::LEVIN "Suegene Levin"                        78 lines  12-DEC-1991 16:23
                          -< NEXPERT OBJECT TRAINING >-
--------------------------------------------------------------------------------
                   ***Announcing NEXPERT Training Course***

Dates:		Oct. 14-18, 1991
		Feb. 17-21, 1992
		June 1-5, 1992

Location:	DLB12, Marlborough, Mass.

Cost:		$1695.00

To register, please contact Suegene Levin, DTN 291-8600, or send the following
information to AIADM::LEVIN -

Name:
Badge No.:
Group Name:
Organization:
DTN:
E-Mail address:
Mailstop:
Manager's Name:
Manager's E-mail:
Cost Center:




Course Description

COURSE:            PROGRAMMING IN NEXPERT OBJECT (TM)

CURRICULUM:        Knowledge Engineering, Emerging Technologies

DESCRIPTION:       Your job is to implement
                   solutions for real problems; problems
                   involving process control, real-time
                   diagnostics and monitoring in manufacturing,
                   design and configuration in electronic and
                   mechanical engineering, or classification
                   and intelligent decision support in finance
                   and insurance. These are real problems you
                   face every day. With instruction based upon
                   real-world experience and hands-on use of
                   NEXPERT Object, you can gain the skills
                   needed to solve real problems with Expert
                   Systems.
                   
GOALS:             At the end of this course, you will be able to:

	           � create a knowledge representation of your
                     problem domain which takes advantage of
                     NEXPERT's architecture and features.

	           � use forward, backward, mixed-mode and
                     context-based inferencing to efficiently
                     control the reasoning process through
                     your knowledge base.

	           � use the NEXPERT callable interface linked
                     to spreadsheets and databases.

	           � ask questions that are most important to
                     your business and develop the answers on
                     a Digital VAXstation.

                   Application programmers,
                   system analysts, knowledge engineers and
                   technical managers who need to know how to
                   use NEXPERT Object to build applications
                   will all benefit from this course.
                   
LENGTH:            5 days

FORMAT:            Lecture/Lab

                                                     Revised June 10, 1991

NEXPERT Object (TM) is a trademark of Neuron Data Inc.
		<<< End of extracted notes >>>

Extraction performed by AISG::ROACH at 04:06 on 17-Dec-91 (Tue)
 using AISG::USER$5:[USER.USER$R.ROACH.NOTES]NOTES.COM;1 -- v37
   with switches "/name=<ROACH_NOTES_WEEKLY> /class=<general_ai> /again_at=<today+7-4:00> /Not_Now=<No>"

67.144Knowledge Based Systems TrainingLMOADM::LEVINMother Nature is a wild womanThu Dec 17 1992 17:34308
				Course announcements
                          Knowledge Engineering Courses
	
Digital's AI Technology Center is offering several courses in knowledge
engineering skills. These courses are open to both Digital employees and Digital
customers. 

These courses enable  a software professional to initiate, select, design, and
program a Knowledge Based System (KBS). Both industry-standard and emerging
technologies and methodologies are studied, while hands-on exercises guide the
participant through the process of knowledge engineering. These courses
culminate in an application workshop in which critical aspects of development
are carried by student teams building a KBS prototype. These courses are offered
both as a 17-day program and may also be taken individually. 
			     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Courses Include

          � Knowledge-Based Systems in Business & Industry       (2 days)
          � Knowledge Acquisition & Representation        (3 days)
          � Knowledge-Based Programming   (5 days)
            Choice of:
            -  NEXPERT Object from Neuron Data Inc.
            -  KBMS from AI CORP
          � Knowledge-Based Prototype/Application Workshop       (7 days)

          *If you do not have a solid programming background, we recommend
          you precede the 17-day course program with DEC OPS5 Programming
          which is a single paradigm, rule-based language.

Target Audience

          Software Engineers, Knowledge Engineers, Domain Experts, and MIS
          professionals interested in developing a repertoire of knowledge
          engineering skills.

Prerequisites

          Students attending the 17-day program are expected to have some
          programming background.

		***************************************
To register for any of these courses, please contact:

Suegene Levin, DTN 296-6600, or complete the registration form at the end of
this announcement and return to LMOADM::LEVIN. 

For customer information and registration, please contact Suegene at (508)
480-6600. 

Individual course descriptions follow.
		***************************************

			
Title: 			Knowledge-Based Systems In Business And Industry

Course code:    	EY-5276E-YC, Reference Code K-SBI

Date:			February 15 - 16, 1993

Location:		LMO2, 111 Locke Drive
			Marlboro, MA

Cost:			Digital internal: $800.00
			Customer external: $1000.00

Length:         	2 Days

Format:         	Lecture/Exercises

Target Audience:        Business Managers,
                        Technical Managers, Software Engineers,
                        Knowledge Engineers, and MIS professionals
                        interested in learning the conceptual
                        foundations for the technology and methodology
                        of Knowledge-Based Systems (KBS) will benefit
                        from this course. It also provides the
                        framework for understanding in-depth techniques
                        of Knowledge Acquisition and Knowledge
                        Representation, and for effectively using KBS
                        programming tools.

Overview:  		This course offers broad
                        coverage of the types and uses of knowledge in
                        organizations, as well as the methods and tools
                        available for preserving, distributing, and/or
                        automating an organization's expertise with KBS
                        technology. Three critical perspectives for
                        success -- business, human resource, and
                        technical -- and are embedded in the course
                        content.

                        Case studies and class exercises enable
                        the practitioner to identify appropriate
                        applications, understand the features of KBS
                        tools, and plan project management and
                        technology transfer.
                          
Prerequisites:        	None


Title: 			Knowledge Acquisition And Representation
                        
Course code:    	EY-5276E-YC, Reference Code K-KAR

Date:			February 17 - 19, 1993

Location:		LMO2, 111 Locke Drive
			Marlboro, MA

Cost:			Digital internal: $1200.00
			Customer external: $1500.00

Length:        		3 Days
                        
Format:        		Lecture/Exercises

Target Audience:        Business Managers,
                        Technical Managers, Software Engineers,
                        Knowledge Engineers, and MIS professionals
                        interested in systematically acquiring and
                        representing knowledge for use Knowledge-Base
                        Systems (KBS) will benefit from this course. It
                        is also recommended for anyone wanting to use
                        KBS programming tools effectively.
                        
Overview:      		This course provides hands-on
                        experience in interviewing managers, experts,
                        and users, analyzing the knowledge acquired,
                        and designing knowledge representations.
                        Specific techniques for knowledge acquisition
                        and representation are presented and practiced
                        independently, followed by an integrated
                        application exercise using video recording and
                        group feedback.

Prerequisites: 		A conceptual foundation in KBS
                        technology and methodology is assumed, as
                        provided in the course "Knowledge-Based Systems
                        in Business and Industry."

Title: 			Programming In NEXPERT Object (tm)
                        
Course code:    	EY-5276E-YC, Reference Code K-NEX

Date:			February 22 - 26, 1993

Location:		LMO2, 111 Locke Drive
			Marlboro, MA

Cost:			Digital internal: $1575.00
			Customer external: $1750.00

Length:        		5 days
                        
Format:        		Lecture/Lab

Audience:      		Software professionals who are
                        interested in a multi-paradigm, mixedchaining
                        KBS development systems for solving knowledge
                        intensive problems in areas such as process
                        control, diagnostics, manufacturing, design and
                        configuration.
                        
Overview:       	NEXPERT Object (tm) provides a
                        variety of important features:  A graphical
                        development environment, a robust reasoning and
                        representation toolkit, and complete
                        integration capability via the Application
                        Programming Interface. Students will learn
                        about and use NEXPERT Object's architecture and
                        features to solve real-world problems involving
                        reasoning,  representation and integration with
                        database and conventional systems.

Prerequisites: 		Students should have knowledge
                        of at least one programming language. Previous
                        experience with knowledge-based systems is
                        helpful.


			NEXPERT Object (tm) is a trademark of Neuron Data


Title: 			Programming in KBMS (R)
                        
Course Code:     	EY-5276E-YC, Reference Code K-KBM

Date:			February 22 - 26, 1993

Location:		LMO2, 111 Locke Drive
			Marlboro, MA

Cost:			Digital internal: $1575.00
			Customer external: $1750.00

Length:        		5 days
                         
Format:        		Lecture/Lab

Target Audience:        Software professionals
                        who are interested in using a multi-paradigm
                        tool with integrated graphical and natural
                        language interfaces.
                        
Overview:        	KBMS (R) (Knowledge Based
                        Management System) is an expert system tool
                        developed by AI Corp. KBMS offers high-level,
                        seamless integration with different database
                        management systems and the ability to be called
                        from existing conventional application
                        programs. The course describes how to develop
                        an expert system using the features of KBMS
                        components  and inferencing strategy.

Prerequisites:          Students should have
                        knowledge of at least one programming
                        language. Previous experience with expert
                        systems is not necessary but may be helpful.

                        KBMS (R) is a registered trademark of AI Corp.


Title: 			Prototyping Knowledge-Based Systems
                        
Course code:    	EY-5276E-YC, Reference Code K-PRT

Date:			March 1 - 9, 1993

Location:		LMO2, 111 Locke Drive
			Marlboro, MA

Cost:			Digital internal: $2200.00
			Customer external: $2450.00

Length:        		7 days
                        
Format:        		Lab

Target Audience:        Software professionals
                        who wish to enhance their KBS development
                        skills using one of several tools, including
          		� VAX OPS5
	          	� KBMS (r)
          		� NEXPERT Object (tm).

Overview:      		This course is designed to
                        support the student through the major phases of
                        KBS development, including knowledge
                        acquisition, system design and implementation.
                        Students will work in small groups at their own
                        pace with the aid of expert practitioners. They
                        will incrementally develop an application which
                        observes key learnings in areas such as
                        architecture flexibility, testing and
                        validation, organizational impact and
                        deployment.
                        
Prerequisite:  		Experience with one of the KBS tools listed above

-------------------------------------------------------------------------------
		Digital Employee enrollment form

Course:
Date:
Cost:

Name:
Badge:
Cost Center:
Is this a Foreign cc?:

Organization:
DTN:
E-Mail address:
Mailstop:
CC Mgr's E-Mail:

--------------------------------------------------------------------------------
			Customer Enrollment Form

Course:
Date:
Cost:

Name:
Address:
City:
State:
Zip:
Mailstop:

Phone:
Fax No.:

Sales rep:
Notes:            (training credits, field cc, etc)


CANCELLATION POLICY:

Class registrations may be cancelled without charge up to ten (10)
business days prior to the course start date. Registrants cancelling
after that period and before the first class day incur a cancellation
fee of 50% of the tuition charge. Cancellation on the first day of
class or a no-show will be charged the full tuition fee. Personnel
substitutions may be made at any time provided they meet the
prerequisites for the course. Digital reserves the right to cancel any 
class due to low enrollment.
67.145Neural Networks and Machine LearningLMOADM::LEVINMother Nature is a wild womanThu Dec 17 1992 17:3682
Title:			Neural Networks and Machine Learning

Course Code:		EY-5276E-YC, Reference Code P-NNT

Date:			February 16 - 18, 1993

Cost:			$1250.00 (Digital employee)
			$1350.00 (external)

Location:		LMO2, 111 Locke Drive, Marlboro, MA

Length:			3 Days

Format:			Lecture/Lab

Target Audience: 	Business Managers, statisticians, scientists, 
			engineers, MIS professionals, or anyone:
			interested in enhancing the use of their business data, 
	    		currently confronted with data analysis problems, 
	    		with case history data on their business.

Overview:		Computer Learning Methods are being used today in 
			industries to enhance the value of information 
			available to them.  Companies collect massive amounts 
			of data in areas ranging from market information to 
			technical process control data.  Transforming that 
			data into knowledge about your business is where 
			computer learning can provide a competitive advantage.

			This course provides concepts, examples, and hands-on 
			exercises which illustrate the application of machine 
			learning techniques to business problems. Several 
			kinds of learning systems are presented and criteria 
			for selecting the appropriate method are examined. 
			Learning systems are also compared to more 
			conventional Knowledge-Based Systems (KBS).

			Computer Learning Methods focuses on Neural Network 
			theory and application.  An introduction to 
			alternative computer learning methods including
			Statistical Pattern Recognition, Decision Tree, 
			and Nearest Neighbor techniques is also provided.

  			Heuristics on building neural networks, representing 
			data, and determining appropriate methods will be 
			reviewed.  Students will learn to identify areas 
			where various learning methods can be applied.

			Students will run experiments and construct Neural 
			Networks using a commercial development tool. 

			Real-world applications will be discussed 
			and demonstrated.

Prerequisites:		Students should have a working knowledge of how 
			computer systems are applied to business problems. 
			Those with a knowledge engineering background will 
			benefit most because they will be able to compare 
			learning systems to KBS.


********* To register, complete the following and return to Suegene Levin,
LMOADM::LEVIN. For information contact Suegene at DTN 296-6600. *****************

                Digital Employee enrollment form

Course: Neural Networks
Date: February 16, 1993
Cost: $1250.00 (Digital internal)
    
Name:
Badge:
Cost Center:
Is this a Foreign cc?:

Organization:
DTN:
E-Mail address:
Mailstop:
CC Mgr's E-Mail:

67.146Enterprise Planning ServiceLMOADM::LEVINMother Nature is a wild womanThu Dec 17 1992 17:37126
			COURSE ANNOUNCEMENT


TITLE:		Enterprise Planning Service
		Introduction, User Training and Model Development Training 
	 
DATES:          February 1 - 5, 1993
		March 8 - 12, 1993
		April 5 - 9, 1993
			 
TIME:           9:00 - 5:00

LOCATION:       LMO2, 111 Locke Drive
		Marlboro, MA

COST:           see below

CANCELLATION:	All cancellations must be made at least ten working
                days before the start of class.

------------------------------------------------------------------------
To Register complete the form at the end of this announcement
and return to LMOADM::LEVIN; To register customers or to receive more course
information contact: Suegene Levin, DTN: 296-6600 (508-480-6600) or
LMOADM::LEVIN 
------------------------------------------------------------------------

DESCRIPTION:	Enterprise Planning Service, formerly know as KARMA 
		(Knowledge-based Advanced Resource Management Architecture) 
		is an innovative approach to executive level support for
  		Strategic Planning.

                Determining business strategy is a critical yet difficult
                process. It involves trying to balance and align conflicting
                forces present in an enterprise and its environment in pursuit
                of enhanced business performance. The enterprises that succeed
                are those which most effectively:

                * leverage the KNOWLEDGE of the relevant informed people in an
                  ongoing planning process;
                * handle the CONFLICT OF APPROACH and view that results from a
                  participative approach to planning, and;
                * play out and analyze SCENARIOS to understand the dynamics and
                  complexities which are both internal and external to the
                  enterprise.

                The Planning Service incorporates an environment which allows
                clients to represent a rich picture of their enterprise
                encompassing company structure, market/product strategy,
                technology impact, human resources, productivity, and financial
                considerations. It also facilitates the representation of the
                diversity of view and approach that may exist in interpreting
                such a business model. When integrated with enterprise
                information, it supports sophisticated scenario generation and
                analysis in the search for viable business futures.

		The course is offered in three sections. You may take the 
		first section on its own, or the first two sections on their
		own or all three. The details on each section are outlined
		below.


		SECTION 1: 	WHAT IS THE ENTERPRISE PLANNING SERVICE

		CONTENTS:	Enterprise Planning Service presentation.
				Demonstration of Generic Business Model.
				High Level Business Benefits of the Service.
		  		Overview of the Methodology.

		LENGTH:		Half day
		AUDIENCE:	Senior Executives, Strategic Planners and 
   				Business Analysts
		PREREQUISITES:  None
		FORMAT:		Lecture
		COST:		Free


		SECTION 2: 	USER TRAINING

		CONTENTS:	Enterprise Planning Service User Training.
				Business Model Script.
				Scenario Analysis and Reporting.
				User Discussion.

		LENGTH:		One and a half days
		AUDIENCE:	Strategic Planners and Business Analysts who 
   				plan to become users of the Service
		PREREQUISITES:  Experience using DECwindows applications
		FORMAT:		Lecture, Lab Exercises and Training Notes
		COST:		$500.00 (Digital internal)
				$900.00 (customers)


		SECTION 3:	METHODOLOGY AND MODEL BUILDING	 

		CONTENTS:	Service Delivery Methodology.
				Business Analysis for Enterprise Planning.
				Model Building.
				Service and System Integration.

		LENGTH:		Three days
		AUDIENCE:	Business Analysts who plan to become  
   				Enterprise Planning modelers 
		PREREQUISITES:  Ability to use an editor (e.g. EDT)
		                Experience using DECwindows applications
		FORMAT:		Lecture, Lab Exercises and Training Notes
		COST:		$1100.00 (Digital internal)
				$1500.00 (customers)

==============================================================================
To register, complete the following and return to LMOADM::LEVIN

Course: Enterprise Planning Service
	Introduction, User Training and Model Development Training
Date:
	
Name:
Badge No.:
Group Name:
Organization:
DTN:
E-Mail address:
Mailstop:
Manager's Name:
Manager's E-mail:
Cost Center: