- From: Frank van Harmelen <Frank.van.Harmelen@cs.vu.nl>
- Date: Tue, 20 Aug 2002 15:22:31 +0200
- To: www-webont-wg@w3.org
pat hayes wrote: > There is a general, rather deep, issue lurking here: different ontologies > will be based on different notions of 'thing', so we will probably > eventually need to reconsider the rather simplistic assumption underlying > both DAML and OWL that there is a single monolithic universe which can be > associated with the entire language. The inferential problem is that in my > ontology (of People in the Arts, say), an existential claim might be false > which is true in your ontology (of People Known to the FBI, say); and > still, I have written my ontology with my notion of 'thing' in mind. If > anyone uses facts from both ontologies, they can get unintended conclusions > that aren't valid in either ontology in isolation. Pat, I agree with your identification of the problematic assumption of a single universe, but I fail to see your objection against deriving conclusions from two ontologies that aren't valid in either ontology in isolation. Deriving such additional conclusions is just what you would expect, no? (it happens when you combine two FOL theories). Frank. ----
Received on Tuesday, 20 August 2002 09:22:37 UTC