W3C home > Mailing lists > Public > public-owl-comments@w3.org > March 2009

Re: [LC response] To Jonathan Rees

From: Jonathan Rees <jar@creativecommons.org>
Date: Fri, 20 Mar 2009 17:20:24 -0400
Cc: <public-owl-comments@w3.org>
Message-Id: <BDA22C13-8C2B-4747-856B-F6C39B40B758@creativecommons.org>
To: Boris Motik <boris.motik@comlab.ox.ac.uk>

On Mar 18, 2009, at 4:08 PM, Boris Motik wrote:

> Dear Jonathan,
>
> Thank you for your comment
>     <http://lists.w3.org/Archives/Public/public-owl-comments/2009Jan/0040.html 
> >
> on the OWL 2 Web Ontology Language last call drafts.
>
> We indeed wanted to say that entities are one of the three syntactic  
> categories,
> and not IRIs. To understand why this is so, consider, for example, the
> ObjectHasValue class expression defined in Section 8.2.3 and the  
> accompanying
> UML diagram shown in Figure 8. The UML association "individual" of  
> the UML class
> "ObjectHasValue" does not point to the UML class "IRI"; instead, it  
> points to
> the UML class "Individual". As shown in Figure 2, the UML class  
> "Individual" is
> a UML subclass of the UML class "Entity". Finally, note that the UML  
> class
> "Entity" in Figure 2 has the UML association "entityIRI" to the UML  
> class "IRI".
> Thus, the Syntax document defines OWL 2 ontologies as consisting of  
> "entities
> identified by IRIs", rather than "IRIs that identify entities". This  
> view is
> reflected in the document's introduction, as well as all the other  
> documents.

I guess it didn't occur to me that OWL would use the words "class,"  
"property," and "individual" at variance with the way they're  
ordinarily used in logic, mathematics, and ordinary language, not as  
related to the domain but merely as syntactic entities. Nor did it  
occur to me that "identifies" would be a relation between an IRI  
(syntactic) and an entity (syntactic), since most of the time is means  
what you call "represent", a relation between syntactic entities and  
domain elements.

Now that I understand all this the document makes much more sense.

> We agree with your comment about "can be thought of as primitive  
> terms", and
> have changed the text slightly.

The new text says:

''Entities'', such as classes, properties, and individuals, are  
identified by IRIs. They define the set of primitive ''terms'' of an  
ontology and can be used to represent the basic elements of the domain  
being described.

This still needs wordsmithing. It says that entities define terms,  
which is nonsense. The entities (or the IRIs) *are* the terms. And  
"the basic elements of the domain" is nonsense - the domain doesn't  
inherently have "basic elements"; rather it is the ontology that  
selects or defines domain elements for "representation" by entities. A  
rewrite is needed here.

> We have also replaced "formal conceptualization" with "formal  
> specification". We
> would prefer not to use "conceptual model" because it contains the  
> word "model",
> which seems to be susceptible to misinterpretation.

The text you have is still not true, in my opinion:

     An OWL 2 ontology is a formal specification of a domain of  
interest.

In what sense can an ontology specify a domain of interest? Ordinarily  
ontologies are descriptive or predictive, not prescriptive. The most  
accurate statement would be

     An OWL 2 ontology is a formal axiomatization of a domain of  
interest.

but I can understand if you think this is too stuffy. One finds  
"formal model" in the literature (in the sense of formalism-as-model- 
of-reality), but that's dissonant with the use of "model" in model  
theory (which has the opposite sense). Elsewhere in this document you  
talk about "description", and this might work:

     An OWL 2 ontology is a formal description of a domain of interest.

> The following URI can be used to inspect the changes introduced in  
> the Syntax
> document in order to address your comments:
>
> http://www.w3.org/2007/OWL/wiki/index.php?title=Syntax&diff=19729&oldid=19723
>
> Please acknowledge receipt of this email to <mailto:public-owl-comments@w3.org 
> >
> (replying to this email should suffice). In your acknowledgment  
> please let us
> know whether or not you are satisfied with the working group's  
> response to your
> comment.

I think these editorial problems need to be fixed.

> Regards,
> Boris Motik
> on behalf of the W3C OWL Working Group
>
Received on Friday, 20 March 2009 21:21:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 20 March 2009 21:21:07 GMT