Re: OWL-DL and linked data

Andrew

I see this as yet another aspect of the fact that the current semantic 
languages pile lacks a way to express the general notion of coreference. 
The URI  a:bar, published properly as linked data with RDF description, 
has the same referent as b:foo, which identifies a plain vanilla HTML 
page, but a stable and authoritative one.  It's been discussed already 
many times that we lack something between owl:sameAs (too strong) and 
rdfs:seeAlso (too weak). Your problem is quite the same. 
rdfs:isDefinedBy limits you to a certain representation framework (RDF).
I was yesterday updating the data at http://www.lingvoj.org, and thought 
again that I had not the proper expressivity to link e.g., 
http://www.lingvoj.org/lang/da to more authoritative but not 
RDF-friendly URIs like 
http://www.ethnologue.com/show_language.asp?code=dan or 
http://www.sil.org/iso639-3/documentation.asp?id=dan (this one seems to 
be down when I write).

We can wish and be proactive towards their publishers to have such 
resources published as linked data, but meanwhile we are stuck.
More and more I think that adding such expressivity to the langauge pile 
is badly needed. We need RDFSemiotics to become operational.
See: http://esw.w3.org/topic/RDFSemiotics

Bernard

Andrew Newman a écrit :
> So I have been spending a lot of time recently trying to work out how
> to use OWL-DL ontologies and linked data properties effectively.  The
> OWL-DL has properties like rdfs:seeAlso and rdfs:isDefinedBy but these
> are annotation properties, which aren't extensible (can't be made
> sub-properties).  There seems to be a hole in these properties - I'd
> like to have something that indicates that something is defined by
> (authoritative) but isn't RDF.  To put it another way, I think what I
> want to do is reuse dc:identifier or something that would indicate
> that the object of a statement (which is a document HTML or PDF)
> defines the subject.
>
> I'd like to reuse them in the same way that you can in RDFS - so if
> you look at RSS 1, FOAF and SIOC, they can all reuse and extend these
> kinds of properties freely.  Annotation properties seem like a good
> idea but they seem to hamper extensibility (but I think I understand
> why they are like that).
>
> I think I understand the usefulness of annotation properties, the
> ideas of the above properties but I have a feeling that the answer is
> obvious and that I just haven't come across the correct answer.
>
>
>
>   

-- 

*Bernard Vatant
*Knowledge Engineering
----------------------------------------------------
*Mondeca**
*3, cité Nollez 75018 Paris France
Web:    www.mondeca.com <http://www.mondeca.com>
----------------------------------------------------
Tel:       +33 (0) 971 488 459
Mail:     bernard.vatant@mondeca.com <mailto:bernard.vatant@mondeca.com>
Blog:    Leçons de Choses <http://mondeca.wordpress.com/>

Received on Tuesday, 10 June 2008 08:40:30 UTC