Re: URIs for Concepts: Best Practices

On Mon, 2004-04-19 at 22:22, David Menendez wrote:
> Miles, AJ (Alistair)  writes:
> 
> > I wanted to consult you all on this matter.  I have agreement from
> > the EEA to publish the GEMET environmental thesaurus in the SKOS/RDF
> > format.  The next step is to work out with them the URIs they wish to
> > assign to their thesaurus and concepts.  I'm not sure what to
> > recommend to them on this matter.
> 
> Dan Brickley's Wordnet vocabulary service[1] at xmlns.com seems like a
> useful model. Essentially, each concept is given a (non-fragmentary) URI
> which, if dereferenced, returns a description of the concept. Mr
> Brickley's system only returns RDF/XML presently, but there's no reason
> it couldn't also return HTML or something else via content negotiation.
> 
> [1] http://xmlns.com/2001/08/wordnet/
> 
> > I thought to use an http:// based URI base (e.g.
> > http://www.eionet.eu.int/GEMET) and then add the id number of each
> > concept (e.g. http://www.eionet.eu.int/GEMET#204).
> 
> That works, but my preference would be for something like
> <http://eionet.eu.int/GEMET/204>. In practice, using a fragment ID means
> that an HTTP request to a term's URI will return nothing or else a
> description of the entire vocabulary, which I'm guessing is pretty
> large.
> 
I think that this practice would certainly work much better with
PSI/PSID constructs than the fragmentary approach - one resource per
concept is probably a best practice that the Published Subjects TC
should recommend.

> > A first question is, is it OK to use http: URIs for concepts?  Sorry
> > to drag this old chestnut up again, but I need some clear answer on
> > best practices for this.  Are we not at all concerned that the same
> > URI may identify both a thesaurus concept and a resolveable network
> > resource (i.e. the file containing the RDF data)?
> 
> It would be confusing for a URI to identify a thesaurus concept and an
> RDF file. The key, as I see it, is the idea that the response to an HTTP
> Get is a representation of the resource, not the resource itself. The
> fact that <http://xmlns.com/wordnet/1.6/Dog> returns an RDF/XML
> document, doesn't mean that it identifies that particular document. If,
> for some reason, you wanted to talk about that RDF/XML document instead
> of the word "Dog", you would need to use a blank node or a different
> URI.
> 
It is certainly true that content negotiation gives you the problem of
talking about the descriptive resource as opposed to the described
thing. That is a strong argument against content negotiation for RDF /
XTM resources. However, there are still two other options:

1) Embed the RDF / TM markup in its XML form. Then use an rdf:ID
attribute or XTM id attribute so that the reference to the RDF/XTM would
be <http://xmlns.com/wordnet/1.6/Dog#foo>

2) Use a profile of XLink to link to the RDF / TM resource that
describes the concept, and make it completely separate. e.g.
<http://xmlns.com/wordnet/1.6/Dog/dog.rdf>

> Not everyone agrees with this position.
> 
I don't think that a position can ever be established which everyone 
will agree with :-)

> > What do you think of info: based URIs for concepts?
> 
> >From an RDF perspective, it's just as good. From a web perspective, it's
> less useful because it can't be dereferenced.

I tend to agree. I tend to consider the use of URIs for subject
identification as being divided into three categories:
1) The URI resolves to the subject being described
2) The URI resolves to a description of the subject being described
3) The URI is used as a pure, unresolvable identifier

I think (2) gives the greatest possibility for interchange of semantics
if the resource addressed by the URI is human-readable - at some point
the processing of semantics has to be transferred from SW machinery to
wet-ware.

Cheers,

Kal
-- 
Kal Ahmed <kal@techquila.com>
techquila

Received on Tuesday, 20 April 2004 03:01:25 UTC