Re: Quick Guide to Publishing a Thesaurus on the Semantic Web

On Fri, Feb 25, 2005 at 03:59:47PM +0000, Leonard Will wrote:
> In message <421F4682.6010704@cs.vu.nl> on Fri, 25 Feb 2005, Mark van 
> Assem <mark@cs.vu.nl> wrote
> >people that are interested in the (Quick) Guide, are interested 
> >because they (a) want to know if there are benefits of an RDF version 
> >of a thesaurus and (b) aren't experts so would like pointers on how to 
> >do conversion. So maybe I have a wrong idea of the intended audience, 
> >what is your view on this?
> 
> I am probably a member of this audience - I know about thesauri, but 
> don't yet know of the benefits of converting them into a more complex 
> format that we use at present.

Leonard,

The way you put this actually illustrates and confirms
the point I was trying to make in my previous posting.

I think of the "conversion" of an existing thesaurus "into"
SKOS as being a process of expressing an existing model in a
Semantic-Web-enabled form.  The explicit, machine-resolvable
statements it makes using angle brackets may be unreadable
(except by geeks), thus visually conveying the impression
that the model itself is somehow more complex.  Seen from
a modeling standpoint, however, a SKOS version should not
normally be any more "complex" than the original.

The way you put it above suggests you might, in effect,
be looking at those complicated angle-brackets and seeing
the result as "a more complex format" in the way a LaTeX or
MS-Word representation are seen as formats more complex than
plain ASCII.  Even if you do not actually see it this way, 
I am sure that many people will see it this way.

If that is the case, then I think the SKOS guidance
materials need to somehow make clear that the conversion is
not something roughly on the level of file formats.  Rather,
it is about the expression of a human-understandable model in
a machine-processable form.  This process may involve some
interpretive tweaking around the edges, but fundamentally
it is about casting the same model in a different form --
not about making it more complicated, and not about putting
it into a different file format.

> To be convinced that it was worthwhile learning how to do conversion, I 
> would need to see the end-user tools and interfaces that will allow 
> people to use thesauri more effectively and to link them to the 
> databases or web resources that are ultimately being sought.

Good point.

> The development of these tools and interfaces may well influence the way 
> in which the thesaurus data is structured - I am a strong believer in 
> systems being driven by the required output, and much of the discussion 
> here up to now has been rather theoretical and abstract, though I know 
> that there has been some discussion of "use cases". Have I missed work 
> that has been done towards providing helpful front ends that will lead 
> naive searchers and help them to gain the advantage of underlying 
> semantic structures?

I'm assuming one could implement any particular application
functionality or interface to a thesaurus using technology
other than SKOS.  The real value of SKOS, as I see it, lies in
the possibilities that open when a thesaurus is expressed in a
generalized, standard form that lends itself to being shared
or cited between applications and merged or cross-referenced
with other thesauri.

Tom

-- 
Dr. Thomas Baker                        Thomas.Baker@izb.fraunhofer.de
Institutszentrum Schloss Birlinghoven         mobile +49-160-9664-2129
Fraunhofer-Gesellschaft                          work +49-30-8109-9027
53754 Sankt Augustin, Germany                    fax +49-2241-144-2352
Personal email: thbaker79@alumni.amherst.edu

Received on Friday, 25 February 2005 16:43:24 UTC