2 ms·
Doesn't description logic mostly boil down to multi-modal logic, which ought to be representable as a fragment of FOL (w/ quantifiers ranging over "possible wor
by zozbot234 3y ago
Doesn't description logic mostly boil down to multi-modal logic, which ought to be representable as a fragment of FOL (w/ quantifiers ranging over "possible worlds")?
Description logic isn't just found in Cyc, either; Semantic Web standards are based on it, for similar reasons - it's key to making general inference computationally tractable.
- stakhanov 3y agoI'm not trying to be dismissive of description logics. (And I'm not dismissive of Lenat and his work, either). A lot of things can fall under that umbrella term. The history of description logic may in fact be just as old as post-syllogism first-order predicate calculus (the syllogism is, of course, far older, dating back to Aristotle). In the Principia Mathematica there's a quantifier that basically means "the", which is incidentally also the most common word in the English language, and that can be thought of as a description logic too. But the perspective of a Mathematician on this is very different from that of an AI systems "practitioner", and CYC seemed to belong more to the latter tradition.