W3C home > Mailing lists > Public > public-swd-wg@w3.org > March 2009

Re: [SKOS] SKOS ontology sanity-check?

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Sun, 15 Mar 2009 16:42:46 +0100
Message-ID: <49BD21F6.6040102@few.vu.nl>
To: Alistair Miles <alistair.miles@zoo.ox.ac.uk>
CC: SWD WG <public-swd-wg@w3.org>
Hi all,

Following the decision we made at last telecon, 
> [NEW] ACTION: Antoine to amend the labels and other annotations in
> skos rdf in consultation w/ sean [recorded in
> http://www.w3.org/2009/03/10-swd-minutes.html#action03]

I have now implemented my own comments [2] on the SKOS XL RDF file [1].

The new version is at [3]. The diff is at [4]. I have also attached my original comments below, and reported on what I have done for each of them.

Comments are of course welcome! I'd like to draw your attention on the latest comment on disjointess of labelling properties, which we might want to address in the next future...



[1] http://www.w3.org/2006/07/SWD/SKOS/reference/20081001/skos-xl.rdf
[2] http://lists.w3.org/Archives/Public/public-swd-wg/2009Mar/0016.html
[3] http://www.w3.org/2006/07/SWD/SKOS/reference/20090315/skos-xl.rdf
[4] http://www.w3.org/2006/07/SWD/SKOS/reference/20090315/diff-skos-xl.txt

> =============== general comments
> *Dublin core namespace:*
> As for the SKOS vocabulary, the new DC namespace 
> (http://purl.org/dc/terms/) is introduced as namespace in the ontology, 
> but only the legacy one (http://purl.org/dc/elements/1.1/) is used. I 
> would suggest to move to the new one.
> Labeling:
> If a decision is made to change the labels of the SKOS file, we should 
> check the labels of XL as well.

Replaced dc: by dct: and removed definition of dc: namespace from rdf:RDF header.

> skos-xl, skosxl or xl?
> In the XL file comments, the "xl:" namespace abbreviation is used. In 
> the reference the vocabulary is coined as "XL", but the namespace is 
> abbrevated as "skosxl:". And the namespace itself is 
> "http://www.w3.org/2008/05/skos-xl". Is there a way to enforce some more 
> coherence? Personally I'd prefer using consistently the "skos-xl" brand, 
> to keep to the actual namespace.

As a minimal improvement I have replaced xl: by skosxl: in the comments, in accordance with the current versions of the Reference and the Primer.

> *Importing standard SKOS in XL?*
> Actually the SKOS reference also defines the axioms that apply to XL 
> constructs refering to "SKOS+XL" data model. This is perfectly fine, but 
> potentially harmful terminologically, related to the previous comment.
> In fact, considering that XL is an extension for SKOS, couldn't we just 
> OWL-import the SKOS ontology in the XL RDF file? That could streamline 
> all our story. Also this could give more specific spec for the (OWL) 
> reasoners that want to use both SKOS and XL at the same time.

http://www.w3.org/2004/02/skos/core is now imported.

> Formal semantics:
> As for my SKOS file review, I would suggest to already give a try at 
> representing the axioms S55, S56 ans S57 in OWL2, to include them in our 
> spec as soon as OWL2 is moved to a more advanced status.

I'll do that in a companion mail.

> Issue date
> For all vocabulary elements, and contrary to what happen for the SKOS 
> file itself, in XL all resources have a
> <dct:issued>2008-12-04</dct:issued>
> Is there a specific reason for this? And why 2008-12-04, which is not 
> not the date the Reference was published? I would suggest to just remove 
> these triples.

I removed them.

> =============== comments on specific parts of the ontology
> labelRelation:
> <rdfs:comment xml:lang="en">xl:labelRelation is not intended to be used 
> directly, but rather as the basis for a design pattern which can be 
> refined for more specific labeling scenarios.</rdfs:comment>
> could be a skos:scopeNote


> prefLabel, altLabel, hiddenLabel:
> I suggest to make these properties disjoint, to mirror what is done for 
> the simple SKOS labelling properties.
> I know that in general the dumbing down of the XL labels will reveal 
> cases when a label is assigned by different means to a concept. But 
> there might be (hopefully very marginal) cases where instances of 
> xl:Label with different lexicalForms may be asserted (or inferred) to be 
> the same. In such cases the inconsistency would not be detected.

This in fact requires first changing the SKOS Reference section that introduces these elements! That's more than editorial work, and we could well add these axioms in the CR period.

> ================================
Received on Sunday, 15 March 2009 15:43:29 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:31:56 UTC