Re: [VM] Scoping Draft with questions to TF members $swbpd

* Thomas Baker <thomas.baker@bi.fhg.de> [2004-09-10 10:13+0200]
> 
> On Tue, Sep 07, 2004 at 12:44:36PM -0700, Uschold, Michael F wrote:
> > As I understand it, your point is that OWL should be used
> > "out of the box" to represent a thesaurus language directly --
> > rather than using OWL first to represent some ad-hoc language
> > of thesaurus relations and then, in turn, using that ad-hoc
> > language to represent the thesaurus.
> > 
> > [MFU] NO NO!  I'm remaining agnoistic. The matter needs looking into.
> > There may be benefits either way. Or there may be clear preferred
> > choice. 
> 
> Mike,
> 
> Have I correctly understood that you mean to say:
> 
>     There are two alternative ways one might use OWL to
>     express a thesaurus: One could use native OWL constructs to
>     represent thesaurus relations.  Or one could use OWL first
>     to represent a language of thesaurus relations and then use
>     that relation language to represent the thesaurus itself.
> 
> If so, I'm thinking the VM Note might state the issue, present
> a few arguments each way, and point off to any available
> sources of emerging solutions.  Does that sound reasonable?

The SKOS work is one such approach. It uses RDF (and bits of OWL I
think) to describe relations like 'broader' that map to the way 
thesaurus-style systems describe the world. Re-modelling a thesaurus as
an ontology is a seriously expensive and challenging effort; it'd be
good to have better  docs to assist those who attempt this, but more
important is to make sure our enthusiasm for OWL modelling doesn't obscure the 
potential value of thesauri-in-RDF approaches...

Maybe Alistair can comment further?

Dan