W3C home > Mailing lists > Public > www-webont-wg@w3.org > December 2001

Re: UPDATE: initial message concerning syntax

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Mon, 24 Dec 2001 09:07:24 -0500
To: jos.deroo.jd@belgium.agfa.com
Cc: www-webont-wg@w3.org
Message-Id: <20011224090724D.pfps@research.bell-labs.com>
From: jos.deroo.jd@belgium.agfa.com
Subject: Re: UPDATE: initial message concerning syntax
Date: Mon, 24 Dec 2001 01:28:33 +0100

> > > > One place to see the problem is in KIF.  The KIF definition
> > > >    http://logic.stanford.edu/kif/dpans.html
> > > > gives an example where a truth predicate can cause problems.
> > >
> > > I see
> > >
> > > > In the DAML+OIL area, if you require the presence of all syntax-like
> > > > constructs in the semantics you can easily end up with structures like
> > > >
> > > >    :_x complementOf :_x .
> > > >
> > > > in all interpretations.  Now consider whether :_x is an instance of :_x.
> > >
> > > I think that complementOf is an irreflexive property
> > > How could we express that in SWOL?
> > >
> > > --
> > > Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
> >
> > SWOL has no such construct.  Even if it did, you could construct a
> > three-element loop which has the same problem.  There are also lots of
> > other constructs that cause similar problems.
> 
> ok, so we better simply say that
> an odd chain of complementOf is a :schemaInconsistency
> i.e. for all :x, :y, :z
> 
>   { :x ont:complementOf :y . :y ont:sameClassAs :x }
>   log:implies { { :x ont:complementOf :y } a :schemaInconsistency } .
> 
>   :x ont:sameClassAs :x .
> 
>   { :x ont:complementOf :y . :y ont:complementOf :z }
>     log:implies { :x ont:sameClassAs :z } .

OK, you can say this, but what does it mean?  log:implies is not defined in
RDF, RDFS, DAML+OIL, or SWOL.  If you are going to write down syntax, you
need to tell your readers how to interpret the syntax.

Maybe you are trying to axiomatize DAML+OIL or SWOL.  OK, that's fine, but,
again, you have to pick some logical formalism to axiomatize it in, and I
don't see any indication of which logical formalism to use here.

[...]

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


Peter F. Patel-Schneider
Bell Labs Research
Received on Monday, 24 December 2001 09:09:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:46 GMT