Re: LANG: owl:ontology

From: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>
Subject: RE: LANG: owl:ontology
Date: Mon, 16 Sep 2002 10:49:16 +0200

> 
> >
> >
> >
> > As this dicsussion moves on, I'm becoming more confused.  Isn't a
> > daml:ontology simply a tag that wraps all the statements that are
> > "in" the
> > ontology?
> >
> 
> No.
> Conventionally the ontology element is the first child of the rdf element,
> and the triples in the ontology are shown in other children of the rdf
> element.
> 
> e.g.
> 
> <rdf:RDF>
>   <daml:Ontology rdf:about="">
>      <daml:version>1.0</daml:version>
>   <daml:Ontology>
>   <daml:Class rdf:ID="Foo" />
>   <!-- ... -->
> </rdf:RDF>
> 
> The "" in the rdf:about refers to the current document. (it is referred to
> in RFC2396 as a same document reference)
> 
> Peter's analysis about documents unfortunately makes a lot of sense, and
> points to a weakness of RDF's triple model that there is no treatment of
> documents.
> 
> Jeremy
> 

Yes.  


If OWL did not have to be RDF, it would be very useful to make an OWL
ontology be precisely an XML element that contained class, property, and
object definitions.  This is more-or-less how the abstract syntax works.
However, the only reasonable way to translate the abstract syntax notion of
an ontology to RDF is to make an OWL ontology be an RDF/XML document, as is
done in DAML+OIL.

peter

Received on Monday, 16 September 2002 06:10:11 UTC