W3C home > Mailing lists > Public > public-esw-thes@w3.org > February 2007

Re: [SKOS] languages and scripts

From: Jakob Voss <jakob.voss@gbv.de>
Date: Wed, 14 Feb 2007 14:56:04 +0100
Message-ID: <45D314F4.4090702@gbv.de>
To: public-esw-thes@w3.org

Bernard Vatant wrote:

> lang:en      rdf:type      lang:PrimaryLanguage
> lang:US      rdf:type      lang:Region
> lang:en-US   skos:broader      lang:en
> lang:en-US   skos:broader      lang:US
> lang:en-US   rdf:type      ???
> Whatever the type of the latter, it's neither Refion nor PrimaryLanguage, 
> and seems strange to have two broader concepts in two different
> classes. No?
>> 2. No differences in classes (this is implied by 1. if you do
>> inferencing)
> I'm not sure I understand this option, and how it is implied by 1.

If you make lang:PrimaryLanguage, lang:Region, and the missing third
class for refined languages all subclasses of skos:Concept then it is
implied that they are also skos:Concept. The second solution is to
directly use skos:Concept without more detailed classes. The drawback of
subclasses is that you need additional inference and logic to handle them.

>> 3. Put regions and scripts in ConceptSchemes of their own
> I've thought about that option also, and really like it also, but have
> the same concern as the above with classes, having "en-US" with two
> broader concepts in two different ConceptSchemes ... and in which
> ConceptScheme itself? Semantics of ConceptScheme is still largely
> underspecified if specified at all, but I would say that
> broader-narrower should generally be internal to a ConceptScheme. Well,
> not sure about that, since OTOH a Concept can belong to several
> ConceptScheme(s). Alistair do you intend to include ConceptScheme in
> your SKOS semantics proposal? Could help here.

I answered this in my answer to Antoine's mail.

Received on Wednesday, 14 February 2007 13:56:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:45:38 UTC