W3C home > Mailing lists > Public > www-webont-wg@w3.org > September 2002

Re: LANG?/SEM?: using resources

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Wed, 18 Sep 2002 19:03:25 -0400 (EDT)
Message-Id: <20020918.190325.65660422.pfps@research.bell-labs.com>
To: hendler@cs.umd.edu
Cc: www-webont-wg@w3.org

From: Jim Hendler <hendler@cs.umd.edu>
Subject: Re: LANG?/SEM?: using resources
Date: Wed, 18 Sep 2002 15:11:02 -0400

> At 11:18 AM -0400 9/18/02, Peter F. Patel-Schneider wrote:
> >From: Jim Hendler <hendler@cs.umd.edu>
> >Subject: Re: LANG?/SEM?: using resources (was Re: LANG: owl:ontology)
> >Date: Wed, 18 Sep 2002 11:06:40 -0400
> >
> >>  >As far as I can see the only viable route is to be able to use resources
> >>  >without committing to anything related to that resource.  To commit to
> >>  >something in some other ontology/document, use imports.
> >>
> >>  If I understand what he said correctly, Peter and I are in complete
> >>  agreement.  [...]
> >>    -JH
> 
> I take it back, I guess I may not have understood what Peter said...
> 
> >Well, perhaps, but what then was the point of your example?  Wouldn't it
> >have been simpler, and much less confusing, to have said
> >
> >[snip]
> 
> >  > This is the case I really care about.  For imports anything that can
> >>  identify and merge graphs makes me happy - for this case, I care that
> >>  we somehow scope what is included.  I would like [URI2 to not include any
> >>  information from URI1].
> >
> >
> >I took the whole point of your example to be that *something* was
> >transferred from URI1 to URI2.
> 
> I don't want nothing transferred, I want only the ones I explicitely 
> mention to be included in the new graph.  I'm still not clear whether 
> this is or is not what you mean (and why I'm confused).

If a mention in URI2 of a resource from URI1 causes causes the entailments
of URI2 to depend on the contents of URI1 then something is being
transferred (or included, or whatever term you want to call this).

> Let me be as specific as possible and please tell me what is included 
> in the graph:
> 
> URI1 contains the following class definition
> 
> <rdfs:Class rdf:ID="Break">
>    <rdfs:subClassOf   URI2:agendaitem>
>    <rdfs:subClassOf>
>      <daml:Restriction>
>        <daml:onProperty rdf:resource="#start"/>
>        <daml:toClass 
> rdf:resource="http://www.w3.org/2000/10/XMLSchema#timeInstant"/>
>      </daml:Restriction>
>    </rdfs:subClassOf>
> </rdfs:Class>
> 
> URI2 contains the following class definitions:
> 
> <rdfs:Class rdf:ID="AgendaItem">
> </rdfs:Class>
> 
> 
> <rdfs:Class rdf:ID="Topic">
>    <rdfs:subClassOf rdf:resource="#AgendaItem"/>
>    </rdfs:subClassOf>
> </rdfs:Class>
> 
> 
> What I expect is that URI1 will entail that Break is in a subclass 
> relation to AgendaItem, but not that Topic is in a subclass relation 
> with AgendaItem.
>
> What I am less clear on (and could go either way) is if instead URI2 said:
> 
> <rdfs:Class rdf:ID="AgendaItem">
>    <rdfs:subClassOf>
>      <daml:Restriction>
>        <daml:onProperty rdf:resource="#uri"/>
>        <daml:toClass 
> rdf:resource="http://www.w3.org/2000/10/XMLSchema#uriReference"/>
>      </daml:Restriction>
> </rdfs:Class>
> 
> whether this means that if Break has a uri resource it must be a 
> uriReference (or however you want to ensure that the #uri 
> field-related triples are included), but I would still expect Break 
> to be a subclassOf agendaItem.
> 
> 
> Perhaps this simple (and real) example will help.
>   -JH

Well it is an example, but now we are back to the previous situation.
Exactly which triples are to be included?  

My position is simple, the mention of a resource from another ontology
brings in *no* triples.  The entailments of URI1 does not depend in any way
on the contents of URI2.

Jeff's position is also simple, the mention of a resource from another
ontology brings in *all* the triples in that ontology.  The entailments of
URI1 are the entailments of the conjunction of URI1 and URI2.

To be viable, any intermediate position needs to precisely
specify how the contents of URI2 affect the entailments of URI1 when URI1
mentions a resource from URI2.

peter
Received on Wednesday, 18 September 2002 19:03:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:52 GMT