- From: Jos De_Roo <jos.deroo@agfa.com>
- Date: Thu, 12 Jun 2003 00:53:57 +0200
- To: Charles.White@networkinference.com
- Cc: "Dan Connolly" <connolly@w3.org>, "Jim Hendler" <hendler@cs.umd.edu>, "Sean Bechhofer" <seanb@cs.man.ac.uk>, "webont" <www-webont-wg@w3.org>, www-webont-wg-request@w3.org
that's also how I understood it
after we added an inference rule like
{: rdfs:fyi :rule9r3. ?R owl:onProperty ?P; owl:someValuesFrom owl:Thing.
?P owl:inverseOf ?Q; rdfs:range ?O. ?Y ?Q ?X; a ?O} => {?X a ?R}.
we were able to prove that the query
?X a <http://eu.org/owl#EuroMP>.
gives
<http://eu.org/owl#Kinnock> a <http://eu.org/owl#EuroMP>
--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
"Charles White"
<Charles.White@networkinfe To: "Dan Connolly" <connolly@w3.org>
rence.com> cc: "Jim Hendler" <hendler@cs.umd.edu>, "Sean Bechhofer"
Sent by: <seanb@cs.man.ac.uk>, "webont" <www-webont-wg@w3.org>
www-webont-wg-request@w3.o Subject: RE: Need a test -- inverse and oneOf
rg
2003-06-11 11:27 PM
The note about querying could be dropped. It is just a note to remind us of
the query and what results it will return.
chas
> -----Original Message-----
> From: Dan Connolly [mailto:connolly@w3.org]
> Sent: 11 June 2003 12:38
> To: Charles White
> Cc: Jim Hendler; Sean Bechhofer; webont
> Subject: RE: Need a test -- inverse and oneOf
>
>
> 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 18:54:24 UTC