RE: Models

Just a small addition to the modelling difficulties. For term-based
models as well as for concept-based models, it is very hard to cope with
"A USE B + C". Is the concept of A actually in the thesaurus or is it
not? It is provided for, yes, by prescribing the combination of two
concepts that are definitely "in". But do you treat A like any other
concept, or does the model have to change? 
 
Traditional print-based thesauri provide inverse entries such as "B  UF+
A". Thesaurus management software that can cope with these is hard to
find. (Yes there are some, but not many).
 
There's a lot to say on the topic of how to manage synthesised (rather
than enumerated) concepts/terms/notations in knowledge organisation
systems, but no time to say it all now...
Stella
 

*****************************************************
Stella Dextre Clarke
Information Consultant
Luke House, West Hendred, Wantage, Oxon, OX12 8RR, UK
Tel: 01235-833-298
Fax: 01235-863-298
SDClarke@LukeHouse.demon.co.uk
*****************************************************



-----Original Message-----
From: public-esw-thes-request@w3.org
[mailto:public-esw-thes-request@w3.org] On Behalf Of Ron Davies
Sent: 01 March 2005 21:00
To: Miles, AJ (Alistair); public-esw-thes@w3.org
Subject: Models 


Al,

Sorry, I had understood that there had been in the past some discussion
of the 'concept-based model' underlying SKOS as opposed to the so-called
'term-based model' underlying the approach of many thesaurus
practitioners, so I assumed that these differences were already known.
The kinds of information elements not covered by the SKOS core model are
things like History note, Source, Definition, Status, Editorial Note,
Date of input and Date last modified for a non-descriptor. In some use
cases, for example in providing a full thesaurus to another institution
for use at the second institution, or to an organization for translation
of the thesaurus, these could be considered as essential data elements
which would mitigate against use of SKOS. But I can certainly understand
that in the context of the Semantic web these might not be considered
important use cases.

In multilingual thesauri there is also the question of non-descriptors
where there is an association between the non-descriptors, i.e. they
describe a concept on the edge of the domain which has been treated as a
non-descriptor not because there is not literary warrant for them to be
considered as a concept in itself. For example, the OECD Macrothesaurus
as developed by the OECD has Leeward Islands an a non-descriptor (though
the concept is clearly defined, is distinct from other related concepts
and can be expressed in several languages) simply because there is not
enough material to justify it as a concept, though when it is used by
the UN Economic Commission for Latin America and the Caribbean the
non-descriptor is switched over to become a descriptor. In these cases,
it may be useful to associate the different non-descriptors together in
a single non-descriptor record (which represent a concept, no matter how
you define concept). I hesitate to mention this case because the
importance that I accord to this use is not shared by all my colleagues,
but it may still be an important point of consideration for developers
of some multilingual thesauri. 

I hope this helps.

Ron

 At 22:03 28/02/2005, Miles, AJ \(Alistair\) wrote:


Ron,
 
Would you mind writing up in short exactly which parts of the BS8723
model are not covered by the SKOS Core model? 
 
Thanks,
 
Al.
 

--- 
Alistair Miles 
Research Associate 
CCLRC - Rutherford Appleton Laboratory 
Building R1 Room 1.60 
Fermi Avenue 
Chilton 
Didcot 
Oxfordshire OX11 0QX 
United Kingdom 
Email:        a.j.miles@rl.ac.uk 
Tel: +44 (0)1235 445440 

-----Original Message----- 

From: public-esw-thes-request@w3.org
[mailto:public-esw-thes-request@w3.org]On Behalf Of Ron Davies 

Sent: 27 February 2005 09:09 

To: public-esw-thes@w3.org; Thomas Baker 

Cc: public-esw-thes@w3.org 

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



At 14:16 25/02/2005, Thomas Baker wrote:




Hmm, it feels to me that the notion of converting between XML 

and RDF automatically opens a can of worms.  For starters, 

what is an "XML version of a thesaurus"?  Such a document 

could presumably take on any number of forms since the document 

models expressible in XML are theoretically quite diverse.

Even assuming that a conversion could be done automatically, there is
also the question of what information might be lost in doing so. The
model underlying SKOS, for instance, does not include all of the
information in the BS 8743 model, which might make it unsuitable for
certain kinds of uses. 



Ron



Ron Davies 

Information and documentation systems consultant 

Av. Baden-Powell 1  Bte 2, 1200 Brussels, Belgium        

Email:  ron(a)rondavies.be 

Tel:    +32 (0)2 770 33 51 

GSM:    +32 (0)484 502 393