Re: SEM: discussions concerning model theory (Re: ADMIN: draft ftf meeting record)

From: "Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
Subject: Re: SEM: discussions concerning model theory (Re: ADMIN: draft ftf   meeting record)
Date: Wed, 17 Jul 2002 14:34:54 +0200

> > >   :large eg:inconsistentWith owl:oneOf .
> >
> > This doesn't follow in any proposal, as eg:inconsistentWith is not defined
> > in any of the proposals.  In any case, what is the intended meaning of
> > eg:inconsistentWith here?
> 
> Well I think that Jeremy has a maybe much better proposal in
> http://lists.w3.org/Archives/Public/www-webont-wg/2002Jul/0101.html
> point 4.3

I agree that Jeremy's proposal covers most of the bases, and provides a
good framework for producing test cases.

> My main concern is that I would like to know WHY
> some stuff is inconsistent (and I believe in a proof
> form of that giving evidence via deduction rules but
> I agree that this is kind of out of scope here)
> and I think that SEM must specify that.

I don't think that it is in the SEM area to specify that ``giving evidence
via deduction rules'' is out of scope.  I don't even think that it is in
the SEM area to even consider this question.

> > > So I thought (and tried sucessfully out with
> > > http://www.agfa.com/w3c/euler/owl-rules)
> > > why not having that
> > >
> > >   :p owl:extension ( ( :s1 :o1 ) ( :s2 :o2 ) ) .
> > >   :q owl:extension ( ( :s2 :o2 ) ( :s1 :o1 ) ) .
> > >   :r owl:extension ( ( :s1 :o1 ) ) .
> > >   :s3 :p :o3 .
> > >
> > > owl-entails
> > >
> > >   :s1 :p :o1 .
> > >   :q owl:samePropertyAs :p .
> > >   :r rdfs:subPropertyOf :q .
> > >   ( :s3 :o3 ) eg:inconsistentWith owl:extension .
> >
> > Well owl:extension is not in any of the proposals so how can this 
> follow?
> 
> Well of course there isn't such an owlproposed:extension
> but then there shouldn't be a owl:oneOf either...
> unless there is something that I miss
> (it was a beg for help Peter).

Huh?  owl:oneOf is already a constructor in the OWL language.  Are you
proposing that there should be a similar constructor for properties?  (Not
that I'm against having such a constructor, but I've never felt the need
for it.  Note that it is possible to get the effect of this construction
already in OWL.)

[...]

> -- ,
> Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

peter

Received on Wednesday, 17 July 2002 09:54:57 UTC