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

Re: SKOS and M Meta-Modelling Layers

From: Dan Brickley <danbri@danbri.org>
Date: Wed, 31 Oct 2007 12:25:38 +0000
Message-ID: <47287442.8030504@danbri.org>
To: mark.melia@computing.dcu.ie
CC: public-esw-thes@w3.org

Mark Melia wrote:
> 
> Hi all,
> 
> I have a question regarding the SKOS definition. I work in the MDA 
> arena, and have been looking at SKOS and how SKOS can be transformed 
> into a modelling language from the meta-modelling technical space. I am 
> trying to establish how the SKOS definition fits into the model 
> definition M layers.
> 
> As far as I can see
> M3 - RDFS Defintion (self-defined)
> M2 - OWL Definition
> M1 - SKOS Definition (OWL T-Box)
> M0 - Domain model in SKOS (OWL A-Box)
> 
> My only problem with this is the fact that the domain model is at the 
> M0, yet in tradition MDA models of this type would be defined at the M1 
> level. Does anyone have any thoughts on this matter?

SKOS is instance-level data from a generic RDF (and RDFS/OWL) 
perspective. But it clearly allows domain modeling, in a simplified 
form. This is classic compsci hack of solving certain problems by adding 
another layer of indirection. Whatever formalities we create for 
carefully structuring data, people will always be able to step down a 
level and do it their own slightly scruffier way. SKOS just documents 
one such common descriptive pattern. So I think this is OK, even if they 
layering seems quirky.

cheers,

Dan
Received on Wednesday, 31 October 2007 12:26:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:38:58 GMT