W3C home > Mailing lists > Public > public-swd-wg@w3.org > February 2007

Re: [SKOS] semantics

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Sun, 11 Feb 2007 19:32:44 +0100
Message-ID: <45CF614C.9010704@few.vu.nl>
To: "Miles, AJ \(Alistair\)" <A.J.Miles@rl.ac.uk>
CC: public-swd-wg@w3.org

Hi Alistair,

Really nice work! The questions you mention are not trivial indeed, but 
I think you document is a great help for our WG task to find an answer 
to them.

Some points on the form however, in my opinion important for future 
discussion:

1. I would suggest that the document makes the distinction between what 
you are proposing for the future version of SKOS and what is assumed by 
the current version [1].
For example, concerning the skos:broader property: you now say that it 
is not transitive by default, which is perfectly fine with me btw. But I 
think you should say that SKOS available specification considers that it 
is transitive. This can influence the debate I suppose... Nothing more 
than one line for each semantic condition (or note) would be needed, I 
suppose.

2. In the same line, I think the semantic links between SKOS constructs 
and RDF(S) and OWL ones should also be mentioned, at least for the 
current SKOS version. E.g that skos:broader is an instance of 
owl:TransitiveProperty. Or that skos:prefLabel is a subProperty of 
rdfs:label.

Cheers,

Antoine

[1] http://www.w3.org/TR/swbp-skos-core-spec/

>Hi all,
>
>There are two basic challenges with respect to the semantics of SKOS.
>
>Firstly, what are the semantics of SKOS?
>
>Secondly, how should the semantics of SKOS be specified?
>
>Niether question is trivial.
>
>The difficulty with the first question is that, while there are some semantic conditions which are intuitively obvious (such as skos:broader and skos:narrower being each other's inverse), there are others which are not at all obvious and for which there may be viable alternatives (such as reflexivity of skos:broader).
>
>The difficulty with the second question is that, even if we only consider those semantic conditions which are not contentious, not all of those conditions may be specified by means of either RDFS or OWL assertions.
>
>One *possible* way forward would be to define the semantics of SKOS as a semantic extension of RDFS, using the definitions and notational conventions laid out in the RDF Semantics [1].
>
>To explore the feasibility of this approach, and to provide a concrete basis for discussion of specific semantic conditions, I have written a draft semantics of SKOS as a semantic extension of RDFS:
>
>[2] http://www.w3.org/2006/07/SWD/wiki/SkosDesign/RdfsSemanticExtension?action=recall&rev=6
>
>Note that this draft only considers a limited subset of the current SKOS vocabulary - but don't worry, there is still plenty of potential for debate :)
>
>[DONE] ACTION: Alistair to summarizes the aspects of semantics of the skos data model [recorded in http://www.w3.org/2007/01/22-swd-minutes.html#action06]
>
>Cheers,
>
>Alistair.
>
>[1] http://www.w3.org/TR/2004/REC-rdf-mt-20040210/
>
>--
>Alistair Miles
>Research Associate
>CCLRC - Rutherford Appleton Laboratory
>Building R1 Room 1.60
>Fermi Avenue
>Chilton
>Didcot
>Oxfordshire OX11 0QX
>United Kingdom
>Web: http://purl.org/net/aliman
>Email: a.j.miles@rl.ac.uk
>Tel: +44 (0)1235 445440
>
>
>
>
>  
>
Received on Sunday, 11 February 2007 18:32:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:28 GMT