W3C home > Mailing lists > Public > www-webont-wg@w3.org > June 2003

RE: Need a test -- inverse and oneOf

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>
Message-Id: <1055360297.2969.5.camel@dirk.dm93.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 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:01 GMT