[ontolex] Public Call for Use Cases (comments from Thomas Baker)

Dear Philipp/all,

I sent the call also to the "DCMI Vocabulary Management Community"
list http://dublincore.org/groups/vocabulary-management/
Thomas Baker sent me some comments on the role of SKOS-XL that I would
like to share with you. Here you are an extract of the thread.

Regards,
Jorge

------
1) 2012/4/2 Thomas Baker wrote:

It's great to see this initiative.  Just a quick note to point out that the
domain for skos:prefLabel is not restricted to SKOS concepts. It can be used to
associate a lexicalization, in principle, with anything, even a property.  The
same holds for the labeling properties in the skosxl: namespace, which are
likewise not restricted with regard to domain.  Indeed, lexicalizations in
multiple languages was one of the main use cases for SKOS-XL.

I'm sure these have come up in discussion... I just wanted to emphasize that
the domain for SKOS labeling properties was deliberately left unrestricted so
that they would be usable, in principle, for labeling anything...

Tom Baker (formerly of W3C Semantic Web Deployment Working Group)

-------
2) 2012/4/2 Jorge Gracia wrote:

Thanks for your email, this is indeed a very good point. My view is
that the model we want to develop in this initiative is for capturing
rich linguistic structures (not only labels, but PoS, variants,
syntactic decompositions, ...) that are not easy to capture with the
existing techniques or not well standardised yet for its consume on
the Semantic Web. But of course SKOS-XL covers already a wide range of
use cases in terms of ontology lexicalization. For this reason we will
have to be able to answer the question "this is already solved by
SKOS-XL?" each time a new problem arises or a new use case is
presented...

------
3) 2012/4/2 Thomas Baker wrote:

Understood!

>                   But of course SKOS-XL covers already a wide range of
> use cases in terms of ontology lexicalization. For this reason we will
> have to be able to answer the question "this is already solved by
> SKOS-XL?" each time a new problem arises or a new use case is
> presented...

Or where the properties and classes of SKOS-XL fall short of meeting
requirements, the group could consider either building an extension on the
basis of SKOS-XL, or perhaps simply borrowing its "literal form" convention
(skosxl:literalForm) in creating new properties [1,2].

Tom

[1] http://www.w3.org/TR/skos-reference/#L5739
[2] http://www.w3.org/TR/skos-reference/#L780

Received on Friday, 13 April 2012 09:14:14 UTC