Re: Representation of Compound Mapping in SKOS

On 6/21/11 4:37 PM, Dan Brickley wrote:
> On 21 June 2011 16:02, Stella Dextre Clarke<stella@lukehouse.org>  wrote:
>> In support of Christine's suggestion I'd point out that the issue has been
>> aired at several events in the past year, e.g. the Cologne Conference  on
>> Interoperability and Semantics in Knowledge Organization
>>   (http://linux2.fbi.fh-koeln.de/cisko2010/presentations.html) and the NKOS
>> workshop in Glasgow
>> (http://www.comp.glam.ac.uk/pages/research/hypermedia/nkos/nkos2010/programme.html),
>> where I gave a presentation on the Types of mapping recommended in ISO
>> 25964.
>>
>> There seems no doubt that for applications in which good quality mappings
>> are needed from one thesaurus to another, there is a demand for compound
>> equivalence mappings. ISO 25964 includes recommendations for the development
>> and use of such mappings. And in another reply to this posting, Johan de
>> Smedt presents some ideas for a schema to use.
>>
>> At both the above conferences it was pointed out that an extension to SKOS
>> would be required. Is there any hope that W3C may soon be able to
>> support/encourage work on such an extension?
>
> I agree that some documentation and/or extensions on how to handle
> this would be very useful.
>
> We have many, perhaps even most, of the right people already here on
> the mailing list. The only real difference that a formal W3C group
> (incubator, working group, whatever category...) adds is in terms of
> having regular structured, chaired discussion and explicitly stated
> goals. If members of this mailing list were to self-organize and make
> a first pass at the problem, that would be a great step forwards. It
> would also help make the case for a more official W3C effort, if one
> were considered necessary.
>
> But whether W3C itself "does something" or not, the real work will be
> done by the experts gathered here. Perhaps having a W3C initiative to
> extend SKOS would make it easier for participants to justify time to
> their projects, line managers, funders etc? Or perhaps it would give
> them stronger confidence that the collaboration would be fruitful. I'm
> not sure. My recommendation is for people to jump in and get started,
> rather than wait for W3C itself to "do something".
>
> Perhaps the Wiki at http://www.w3.org/2001/sw/wiki/SKOS is a good hub
> for collaborations? Use cases, examples, links to draft specs etc?
>



We had that discussion a couple of times in the past, already :-)

I think yes, it really has to be bottom-up. At that stage W3C can probably encourage it--extending ontologies is always a good practice! But concrete support will be more difficult right now.
And, yes, the people on this list should have enough expertise to do something...

Note that the SKOS issue page (http://www.w3.org/2001/sw/wiki/SKOS/Issues) still identify these issues, starting from the moments they were identified by the W3C group. But it never really took up. There were some proposals around, yet, which could still be useful to have a look at. They may overlap quite a bit with DNB's approach or Johan's ontology.
In particular, two issues are relevant:
-http://www.w3.org/2001/sw/wiki/SKOS/Issues/BooleanExpressionMappings
- http://www.w3.org/2001/sw/wiki/SKOS/Issues/Coordination (and yes, I know that the problem may be theoretically different, but any pattern used to solve it could be re-used for the previous issue!)

Cheers,

Antoine

Received on Tuesday, 21 June 2011 15:03:54 UTC