T.R | Title | User | Personal Name | Date | Lines |
---|
1627.1 | I can link those words in ... | VMSDEV::HALLYB | Fish have no concept of fire. | Fri Jun 12 1992 14:35 | 2 |
| 13 steps, if you permit �fils� which is in my dictionary as
1/1000 of a dinar.
|
1627.2 | | TRACE::GILBERT | Ownership Obligates | Fri Jun 12 1992 18:32 | 3 |
| �fils� wasn't in my dictionary (which was the on-line Scrabble � dictionary.
But I found it in the new Webster's Unabridged -- with it's French definition
and no mention of dinars!
|
1627.3 | 10 here | WECARE::GRIFFIN | | Sat Jun 13 1992 02:22 | 8 |
| I get the transformation in 10 steps, using non-obscure words.
Isn't the general problem here one of language (or word) recognition
--- as in, "automata theory"? From the complexity point of view, if
we're talking about checking transformations against a finite
dictionary, (given various relevant restrictions in place on the
English language), I don't think this is a "hard" problem.
|
1627.4 | Suggestions for part 3 | UNTADA::TOWERS | | Mon Jun 15 1992 08:38 | 13 |
| Apart from length I'd say that were two other obvious metrics for
difficulty -
Choice
The more choices you have at any stage the more difficult the
problem. Hence the example in .0 is easy for its length because
there are few choices for where to go from INKS.
Obscurity
The more words along the way that you have to get right that are
obscure, the more difficult the problem.
Brian
|
1627.5 | optimality? | MOCA::BELDIN_R | All's well that ends | Mon Jun 15 1992 10:15 | 7 |
| -.1 suggests that too many choices and too few are both likely to
complicate the problem, "too many" because of combinatorial explosion
and "too few" because of the "recognition" problem (which depends on
the efficiency of the search technique). That combination suggests
there may be some way to formulate an optimality criterion.
/rab
|
1627.6 | x-ref | DESIR::BUCHANAN | | Mon Jul 06 1992 06:17 | 5 |
| > Are there word linkages that require more steps?
See note 1565.
Andrew.
|