- From: Dan Connolly <connolly@w3.org>
- Date: 11 Jun 2003 14:38:17 -0500
- To: Charles White <Charles.White@networkinference.com>
- Cc: Jim Hendler <hendler@cs.umd.edu>, Sean Bechhofer <seanb@cs.man.ac.uk>, webont <www-webont-wg@w3.org>
On Wed, 2003-06-11 at 13:31, Charles White wrote: > All, > > I just sent a message to Chris and Jim about a combined test. Here is one we have put together at Network Inference, and that we can handle with our engine. Many thanks for the detail work... > <rdf:RDF > xmlns:my="http://eu.org/owl#" pls change to example.org. > xmlns:owl="http://www.w3.org/2002/07/owl#" > xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" > xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" > xml:base="http://eu.org/owl" I expect the test editor will tweak that line too. > > > > <owl:Class rdf:ID="EuropeanCountry" /> > <owl:Class rdf:ID="Person" /> > > <owl:Class rdf:ID="EUCountry"> > <owl:oneOf rdf:parseType="Collection"> > <my:EuropeanCountry rdf:ID="UK"/> > <my:EuropeanCountry rdf:ID="BE"/> > <my:EuropeanCountry rdf:ID="ES"/> > <my:EuropeanCountry rdf:ID="FR"/> > <my:EuropeanCountry rdf:ID="NL"/> > <my:EuropeanCountry rdf:ID="PT"/> > </owl:oneOf> > </owl:Class> > > <owl:ObjectProperty rdf:ID="hasEuroMP" > > <rdfs:domain rdf:resource="#EUCountry"/> > </owl:ObjectProperty> > > <owl:ObjectProperty rdf:ID="isEuroMPFrom" > > <owl:inverseOf rdf:resource="#hasEuroMP"/> > </owl:ObjectProperty> > > <owl:Class rdf:ID="EuroMP"> > <owl:equivalentClass> > <owl:Restriction> > <owl:onProperty rdf:resource="#isEuroMPFrom" /> > <owl:someValuesFrom rdf:resource="http://www.w3.org/2002/07/owl#Thing" /> > </owl:Restriction> > </owl:equivalentClass> > </owl:Class> > > <my:Person rdf:ID="Kinnock" /> > > <my:EuropeanCountry rdf:about="#UK"> > <my:hasEuroMP rdf:resource="#Kinnock" /> > </my:EuropeanCountry> > > </rdf:RDF> > > <!--querying for all subconcepts of EuroMP should return "Kinnock" --> I don't quite understand that last line. Is this still an inconsistency test? > > -----Original Message----- > > From: Jim Hendler [mailto:hendler@cs.umd.edu] > > Sent: 11 June 2003 05:20 > > To: Sean Bechhofer > > Cc: webont > > Subject: Re: Need a test -- inverse and oneOf > > > > > > > > That would serve my needs. Jeremy.could we turn this into a proposed > > incocnsistency test? > > > > At 1:14 PM +0100 6/11/03, Sean Bechhofer wrote: > > >On Wed, 11 Jun 2003, Jim Hendler wrote: > > > > > >> > > >> To be able to close some of our LC comments, there must > > be a test in > > >> our test suite that includes both inverse and oneOf. In > > an earlier > > >> message I outlined a simple one, but perhaps someone in the WG can > > >> come up with a better one. It is my opinion that without > > such a test > > >> we will have trouble convincing some people that OWL DL is > > >> implementable (cf the comments from Jena and HP). Can we > > please get > > >> such a test at least proposed?? > > >> thanks > > >> JH > > > > > >How about the following? > > > > > >Ontology( > > > > > > Class(a:NiceCorporation partial > > > restriction(a:employs allValuesFrom a:NiceGuy) > > > restriction(a:employs someValuesFrom oneOf(a:tom a:dick))) > > > > > > Class(a:NiceGuy) > > > > > > ObjectProperty(a:employedBy > > > inverseOf(a:employs)) > > > ObjectProperty(a:employs) > > > > > > Individual(a:dick > > > type(complementOf(a:NiceGuy))) > > > > > > Individual(a:niceCorp > > > type(a:NiceCorporation)) > > > > > > Individual(a:tom > > > type(restriction(a:employedBy allValuesFrom > > >complementOf(a:NiceCorporation)))) > > > > > >) > > > > > >The interaction of the oneof and the assertion that dick > > isn't a Nice Guy > > >allows us to conclude that niceCorp must employ tom. But > > then we know that > > >anything that employs tom cannot be a NiceCorporation (due to the > > >inverse), so we get an inconsistency. > > > > > >It's pretty trivial, but I think you do need both one-of and > > inverse to be > > >able to state it. > > > > > > Sean > > > > > > -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Wednesday, 11 June 2003 15:38:02 UTC