Re: SKOS-XL (was RE: SKOS/ synonym provenance (ISSUE-27 AnnotationOnLabel))

I also have a perceived need for something like SKOS-XL in the absence
of labels-as-resources. The specific problem I have is dealing with
pre-coordinated labels in Library of Congress Subject Headings (LCSH).
For example when modeling a heading of "Logic, Modern--19th century"
I'd like to have the complete string as the prefLabel while also being
able to model the fact that "Logic, Modern" is a topical component and
"19th Century" is a chronological component using my own (not yet
invented) vocabulary. Unfortunately not all LCSH subdivisions like
"19th Century" can stand alone as Concepts, so they can't really be
post-coordinated as in MESH. So something like:

lcsh:123 rdf:type skos:Concept;
  skos-xl:prefLabel lcsh:456

lcsh:456 rdf:type skos-xl:Label;
  skos-xl:plainLiteralForm "Logic, Modern--19th Century"@en.

lcsh:456 rdf:type lcsh:SubjectHeading
  ...

I appreciate the desire to keep SKOS lightweight, but I also want to
amplify Alan's point about the convention being widely adopted amongst
SKOS users. I've been privately hopeful that labels-as-resources
hadn't been entirely ruled out since ISSUE-26 [2] and ISSUE-27 [3] are
still open and there is a candidate requirement in the use cases for
R-AnnotationOnLabel. But I'm also open to exploring the practical side
of something like SKOS-XL. As a newbie I'm particularly interested in
how the entailment rules would get expressed, and processed in the
"real world".

Is it worthwhile at this point for interested folks to write up
SKOS-XL as a Note? I imagine it would need to take a back seat to the
Reference and Primer work... My personal opinion is that I think it
will be important for the Primer to directly address extensibility of
labels.

//Ed

[1] http://www.w3.org/2006/07/SWD/track/issues/26
[2] http://www.w3.org/2006/07/SWD/track/issues/27

Received on Wednesday, 21 November 2007 15:00:21 UTC