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

RE: Need a test -- inverse and oneOf

From: Jim Hendler <hendler@cs.umd.edu>
Date: Wed, 11 Jun 2003 19:16:30 -0400
Message-Id: <p05200f07bb0d68887403@[10.0.1.2]>
To: "Jos De_Roo" <jos.deroo@agfa.com>, Charles.White@networkinference.com
Cc: "Dan Connolly" <connolly@w3.org>, "Sean Bechhofer" <seanb@cs.man.ac.uk>, "webont" <www-webont-wg@w3.org>, www-webont-wg-request@w3.org

Boy - I wish all my requests to this WG were handled this fast and 
this well!!  Many thanks to Sean, Charles and Jos -- I believe the 
fact that we have a couple of tests, and some implementations that 
can handle them, will help in convincing people that OWL DL can, 
indeed, be implemented!
  -JH



At 12:53 AM +0200 6/12/03, Jos De_Roo wrote:
>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/
>>
>>
>>

-- 
Professor James Hendler				  hendler@cs.umd.edu
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  *** 240-277-3388 (Cell)
http://www.cs.umd.edu/users/hendler      *** NOTE CHANGED CELL NUMBER ***
Received on Wednesday, 11 June 2003 19:16:50 GMT

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