- From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
- Date: Sat, 19 Apr 2003 08:28:51 -0400 (EDT)
- To: phayes@ai.uwf.edu
- Cc: w3c-rdfcore-wg@w3.org
Arrgh. Sorry. You're right. I was confused. I'm not used to thinking of restrictions as things that need to be considered as classes. peter From: pat hayes <phayes@ai.uwf.edu> Subject: Re: iff Date: Fri, 18 Apr 2003 16:38:38 -0500 > >Hmm. Under this reasoning > > > >ex:foo rdf:type owl:Restriction . > >ex:foo owl:onProperty rdf:type . > >ex:foo owl:hasValue rdfs:Class . > > > >implies > > > >ex:foo rdfs:subClassOf rdfs:Class. > > > >I don't think so. > > I do. The argument: suppose (x type foo); then (x type Class) > because of the restriction on type; but x is arbitrary, so (foo > subClassOf Class) by the iff condition on rdfs:subClassOf. > > And this seems to me to make intuitive sense, since the restriction > amounts to saying that anything that foo is true of must be of type > rdfs:Class, so foo is a subclass of rdfs:Class. Obviously it would > work for any class, not just rdfs:Class. Also in OWL-Full it would > work with ranges as well, since they also have iff semantics. > > Pat > > PS the logical translation of the above amounts to > foo(x) implies type(x,Class) > which with the axiom > type(x,y) iff x(y) > gives > foo(x) implies Class(x) > with x free, ie universally quantified; which is a sufficient condition for > Subclass(foo, Class) > because > Subclass(x,y) iff (all (z)(x(z) implies y(z) ) > > > > > -- > --------------------------------------------------------------------- > IHMC (850)434 8903 or (650)494 3973 home > 40 South Alcaniz St. (850)202 4416 office > Pensacola (850)202 4440 fax > FL 32501 (850)291 0667 cell > phayes@ai.uwf.edu http://www.coginst.uwf.edu/~phayes > s.pam@ai.uwf.edu for spam > >
Received on Saturday, 19 April 2003 08:29:05 UTC