Semantic Web intelligence

It's a half-baked, non-academic idea, but it just occurred to me: the Semantic Web is essentially an architecture for distributed knowledge representation. As such, it's quite a good element of some intelligent, subsumption-based agent — the system can farm out knowledge storing and inference, and focus on (a) discovery of domain-specific rules; (b) formulation of the right queries; (c) investigation of knowledge sources.

This seems an interesting approach to me; intelligently examining query results to determine additional procedures for knowledge generation, and finding out what questions to ask and where to get the data to answer them.

I agree with Patrick Hayes that the importance of URIs to the Semantic Web is often underestimated; their possibilities for distributed referencing and the interaction with the SW's assumptions are quite special.

Posted at 2004-05-21 15:30:42 by Richard • Link to Semantic Web intel…
Comments, trackbacks.