Re: [SKOS] About Closing ISSUE-36 ConceptSchemeContainment

Hello,

Regarding my action on ISSUE-36 [1]. It seems that my mail [4] was a bit 
too panicky.
As a reminder, the complete text of ISSUE-36 reads:

> SKOS defines a 'concept scheme' as: "a set of concepts, optionally including
> statements about semantic relationships between those concepts."
>
> SKOS relationships such as broader and narrower are represented as triples in
> RDF. The fact that a particular broader/narrower relationship between two
> concepts belongs to a concept scheme cannot then be represented without
> resorting to reification.
>
> A principled approach to representing this containment would be desirable.


One can indeed select/adapt from [2, 3, 4] the following material:

=============== Beginning of text

Vocabulary: skos:ConceptScheme, skos:inScheme, skos:hasTopConcept

Axiomatic Triples:
skos:ConceptScheme rdf:type owl:Class.
skos:inScheme rdf:type owl:ObjectProperty
skos:inScheme rdfs:range skos:ConceptScheme
skos:hasTopConcept rdf:type owl:ObjectProperty
skos:hasTopConcept rdfs:domain skos:ConceptScheme
skos:hasTopConcept rdfs:range skos:Concept
skos:ConceptScheme owl:disjointWith skos:Concept

skos:ConceptScheme denotes the class of SKOS concept schemes. Beyond 
this statement, there are no further semantics conditions on the 
interpretation of skos:ConceptScheme.

This specification does not make any statement about the formal 
relationship between the class of Concept Schemes and the class of Named 
RDF Graphs. The decision not to make any such statement has been made to 
allow different design patterns to be explored for using SKOS with query 
languages such as SPARQL. @@For more information about recommended 
patterns for using SKOS with SPARQL, see SECTION@@
In particular, skos:ConceptScheme MAY be interpreted as a sub-class of 
the class of named RDF graphs. This would allow to use the name (URI) of 
a concept scheme in SPARQL queries as the name of a graph, to establish 
the containment in this concept scheme for a semantic relationship 
between two SKOS conceptual resources.
Notice that this interpretation would not be appropriate, however, if 
different named RDF graphs were used to express different "states" or 
"versions" of a concept scheme; or if a concept scheme were interpreted 
as having alternative expressions, as an RDF graph and an HTML document 
for example (in which case separate URIs might be required for the 
concept scheme, the RDF graph, and the HTML document).

skos:ConceptScheme MAY also be interpreted as a sub-class of 
owl:Ontology. This would be consistent with using owl:imports to make 
logical import statements between SKOS concept schemes.

It is also possible to use rdfs:isDefinedBy to explicitly state the 
relationship between a SKOS conceptual resource and the concept scheme 
in which it is defined.
However, for the purpose of stating the relationship between a SKOS 
conceptual resource and the concept scheme(s) to which it belongs, which 
is a different goal, the skos:inScheme property shall be used.

=============== End of text

I think this gives an answer to ISSUE-36.
Notice that I've made is to replace "provenance" by "containment" in the 
following sentence from [2]
>
> In particular, skos:ConceptScheme MAY be interpreted as a sub-class of 
> the class of named RDF graphs. This would allow to use the name (URI) 
> of a concept scheme in SPARQL queries as the name of a graph, to 
> establish the provenance of a semantic relationship between two SKOS 
> conceptual resources.

I've done this to better fit the ISSUE. Please say if this has 
consequences I have overlooked...

Best,

Antoine

[1] http://www.w3.org/2006/07/SWD/track/issues/36
[2] 
http://www.w3.org/2006/07/SWD/wiki/SkosDesign/ConceptSchemes/MinimalProposal?action=recall&rev=1
[3] http://www.w3.org/2006/07/SWD/SKOS/reference/20071223
[4] http://lists.w3.org/Archives/Public/public-swd-wg/2008Jan/0019.html

Received on Tuesday, 15 January 2008 20:57:12 UTC