proposed partial reply for [Re: Remarks on OWL Guide and question about AS&S]

I propose the following reply for part of

> Concerning OWL AS & S
> I have not clearly understood the goal of section 4.1 (I must recognize I am
> far from being an expert in those matters). If it is to obtain (when
> possible) equivalent RDF triples for OWL definitions, why not using RDF
> abilities at best ? In particular, could an equivalent for
> "EquivalentClasses (D1...Dn)" be
> "T(Di) rdfs:subclassOf T(Di+1); T(Di+1) rdfs:subClassOf T(Di),1<=i<n" or
> would a RDFS-guru find it really stupid ?

This would work as a mapping.  However, the intent of the mapping is not to
translate into something that can be partly understood by an RDF or RDFS
processor, but instead is to provide a mapping between the abstract syntax
and the exchange syntax to define which RDF graphs belong to the OWL DL and
OWL Lite dialects of OWL.  The mapping you propose above would not work
quite as well for this purpose.

I will add the following paragraph (or something close to it) to the
beginning of Section 4 to make some of this clearer.

<span class="change">
This section of the document provides a mapping from the abstract syntax
for OWL DL and OWL Lite given in <a href="syntax.html">Section 2</a>
to the exchange syntax for OWL, namely 
RDF/XML [<cite><a href="#ref-rdfsyntax">RDF Syntax</a></cite>].
This mapping (and its inverse) provide the normative relationship between
the abstract syntax and the exchange syntax.
It is shown in <a href="rdfs.html">Section 5</a> and 
<a href="proofs.html#A.1">Appendix A.1</a> that this mapping preserves the
meaning of OWL DL ontologies.
<a href="#4.2">Section 4.2</a> defines 
the OWL DL and OWL Lite dialects of OWL as those RDF graphs
that are the result of mappings from abstract syntax ontologies.

Does this appropriately answer your query?

Peter F. Patel-Schneider 
Bell Labs Research
Lucent Technologies

Received on Wednesday, 14 May 2003 10:27:08 UTC