W3C home > Mailing lists > Public > public-swd-wg@w3.org > January 2008

Re: [SKOS] About Closing ISSUE-36 ConceptSchemeContainment

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Sat, 19 Jan 2008 19:39:50 +0100
Message-ID: <479243F6.3040108@few.vu.nl>
To: "Miles, AJ \(Alistair\)" <A.J.Miles@rl.ac.uk>
CC: SWD WG <public-swd-wg@w3.org>

Hi Alistair,

Thanks! The new Appendix E and the way it is referred to in Section 4.6.3 
generally makes a nice complement to the existing resolution on 
ISSUE-36. I would support a resolution stating that the issue is closed 
by the mail on skos:inScheme [2] and this appendix.

And I have no problems having my mail not be re-used as such. Actually for the MAY words, they were here because you had put them in the original version of the Reference.

Also:
- for the TODO: for the moment I don't see more patterns that would be 
needed. Were you thinking of different kind of relations (documentary 
notes for instance)?

- with a bit more time to think about it, I have some comments regarding the remark

> Note, however, that this pattern would not be appropriate if different 
> named RDF graphs were used to express different "states" or "versions" 
> of a concept scheme; or if a concept scheme were viewed as having 
> alternate 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).

First, I don't really see what you meant by "states".
Second, I don't see the problem for versions. Indeed, if a version is 
splitted in several graphs (let's say one for the old base version of the concept scheme and a 
new one for additions to it), I would suggest that one 
"main graph" would always be created, and include the information 
contained in other graphs by means of "owl:imports" statements. This 
could solve the issue you mention. It would as well as create a unique 
reference for the "complete concept scheme at version X", which is I 
guess needed anyway for pointing at a given version of a concept scheme.

Cheers,

Antoine

[2] http://lists.w3.org/Archives/Public/public-swd-wg/2007Dec/0058.html


> Dear Antoine,
>
> A new Editors' Draft of the SKOS Reference is available at:
>
> [1] <http://www.w3.org/2006/07/SWD/SKOS/reference/20080118>
>
> This draft includes a new appendix E as a placeholder for SKOS/SPARQL patterns. 
>
> The preamble to this appendix states the following:
>
> """This section describes some patterns for using the SPARQL query language [@@REF-SPARQL-QUERY] to implement some common operations required by applications that use SKOS data. All of these patterns are consistent with the SKOS data model."""
>  
> A sketch has been included of a pattern in which names of concept schemes are used as names of RDF graphs, allowing the containment of a semantic relation to be queries, with the caveat that this might not be appropriate for more advanced versioning scenarios.
>
> Section 4.6.3. in the main body of the document provides a brief statement of our position wrt concept schemes and named RDF graphs, with a link to more detailed information in the new appendix E. 
>
> The originally proposed text (below) has not been used verbatim, because the document does not define a formal notion of conformance (see section 1.7), and therefore the keywords MAY and SHALL are hard to interpret -- it is not used anywhere else in the document (none of the BCP 14 key words are used).
>
> However, appendix E hopefully makes it clear that all of the documented SKOS/SPARQL patterns are consistent with the SKOS data model.
>
> Note also that [1] does not make any statement about recommended practice wrt rdfs:isDefinedBy. If that is considered within scope for this document, then I suggest we raise an issue specifically devoted to that and consider it for subsequent drafts.
>
> I hope this addresses your concerns.
>
> Kind regards,
>
> Alistair.
> 	
> --
> Alistair Miles
> Research Associate
> Science and Technology Facilities Council
> Rutherford Appleton Laboratory
> Harwell Science and Innovation Campus
> Didcot
> Oxfordshire OX11 0QX
> United Kingdom
> Web: http://purl.org/net/aliman
> Email: a.j.miles@rl.ac.uk
> Tel: +44 (0)1235 445440  
>
>   
>> -----Original Message-----
>> From: public-swd-wg-request@w3.org 
>> [mailto:public-swd-wg-request@w3.org] On Behalf Of Antoine Isaac
>> Sent: 15 January 2008 20:57
>> To: SWD WG
>> Subject: 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/M
>> inimalProposal?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 Saturday, 19 January 2008 18:40:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 19 January 2008 18:40:14 GMT