- From: Jim Hendler <hendler@cs.umd.edu>
- Date: Fri, 26 Dec 2003 12:05:25 -0500
- To: Drew McDermott <drew.mcdermott@yale.edu>, public-sws-ig@w3.org, www-rdf-logic@w3.org
At 18:08 -0500 12/19/03, Drew McDermott wrote: >[Cross-posting to try to move this discussion away from the Semantic >Web Services mailing list. Granted it's relevant, but its true scope >is wider.] > > [Frank McCabe] > The 'problem' I was referring to was that of automatically mapping one > ontology (written I assume by person or persons A) to another (written > by persons B). > > People have asserted that there exist automatic tools for doing that. > And I was pointing out some corner cases. > >People are kidding themselves. Not to offend their fans, but the >results on automatic ontology mapping using probabilistic methods, >machine-learning techniques, graph isomorphism, word repetition, >etc. etc. are publishable and will make some of us into academic >stars, but they won't actually begin to solve the problem. The >problem is AI-complete. It's like Richard Waldinger's solution to the >automatic-programming problem: A. Create an intelligent robot. >B. Give it a Lisp manual to read. > > >-- > -- Drew McDermott > Yale Computer Science Department Drew,. I agree completely if we use your definition of ontology-merging. Partial mappings have a greater success (particularly allowing heuristic mechanisms), and of course there's no reason we can't have some human in the loop. Also, none of the literature I know allows instances to be mapped against multiple ontologies, which is a new idea that occurs easily on the Semantic Web, and which opens many opportunities for new research. So I guess I'm kidding myself -JH -- Professor James Hendler http://www.cs.umd.edu/users/hendler Director, Semantic Web and Agent Technologies 301-405-2696 Maryland Information and Network Dynamics Lab. 301-405-6707 (Fax) Univ of Maryland, College Park, MD 20742 240-277-3388 (Cell)
Received on Friday, 26 December 2003 12:07:01 UTC