Re: TEST: scope

On June 29, Peter F. Patel-Schneider writes:
> 
> From: "Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
> Subject: Re: TEST: scope
> Date: Sat, 29 Jun 2002 14:55:50 +0200
> 
> [...]
> 
> > Jim, I agree and I've tried to do that for
> > owl:TransitiveProperty and owl:unionOf in the
> > test repository http://www.w3.org/2002/03owlt/
> > 
> > The gist is that
> > ( <http://www.w3.org/2002/03owlt/pathTransitiveP.n3>
> >   <http://www.w3.org/2001/10/daml+oil> )
> >   log:entails
> >   <http://www.w3.org/2002/03owlt/pathTransitiveC.n3> .
> > 
> > ( <http://www.w3.org/2002/03owlt/unionOfP.n3>
> >   <http://www.w3.org/2001/10/daml+oil> )
> >   log:entails
> >   <http://www.w3.org/2002/03owlt/unionOfC.n3> .
> > 
> > This is using an RDF/N3 presentation syntax
> > but the RDF/XML resources are there as well
> > (I hope with not too many bugs, as I
> > had to do the RDF list stuff by hand
> > instead of using CWM's N3-to-RDF)
> 
> Now wait just a minute here.  Are you actually suggesting that OWL tests
> use N3 or log:entails?  If so, I protest in the strongest terms.  I am
> *not* willing to have any OWL tests be written in a system that has neither
> syntax nor semantics, like N3.  Neither am i willing to have any OWL tests
> written using connectives that do not have a semantics, like log:entails.

I agree. What is wrong with producing formal tests that mirror the
informal examples introduced in the walkthrough/feature-synopsis
document? E.g., for transitive properties, the test could check that:

"if ancestor is stated to be transitive, and if Sara is an ancestor of
Louise (i.e., (Sara,Louise) is an instance of ancestor) and Louise is
an ancestor of Deborah (i.e., (Louise,Deborah) is an instance of
ancestor), then a reasoner may deduce that Sara is an ancestor of
Deborah (i.e., (Sara,Deborah) is an instance of ancestor"

Such a test can easily be formulated in OWL, e.g., given a KB
containing the relevant axioms for the above facts, adding an axiom
stating that Sara is not an ancestor of Deborah should lead to an
inconsistent KB (which is the same as saying that Sara being an
ancestor of Deborah is "owl-entailed" by the KB).

By the way, asserting that Sara is not an ancestor of Deborah can be
done by asserting that Sara is an instance of the class:

     ComplementOf(restriction(ancestor value=Individual(Sara)))

Ian

> 
> peter
>

Received on Sunday, 30 June 2002 01:35:30 UTC