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

Re: ISSUE-146: Last Call Comment: broadMatch and skos:narrowMatch should be used only when there are no exact or close matches for the term elsewhere?

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Wed, 22 Oct 2008 21:04:37 +0200
Message-ID: <48FF7945.7000407@few.vu.nl>
To: Alistair Miles <alistair.miles@zoo.ox.ac.uk>
CC: public-swd-wg@w3.org

+1 for the first part. Disjointness of properties seems for the moment a 
bit difficult to state, given the meagre amount of expertise in that 
domain. Actually for some experiments we have thought about that, and 
came with no definitive conclusion.  It depends on what you want to do 
with the mappings, it seems. Some cases would accomodate very well 
non-disjoint properties, some will actually exploit the disjointess to 
make inferences wrt. to the quality of an alignment.

For the second part, I do not really understand the comment. Where have 
we stated that the semantic relationships are of secondary importance? I 
mean, having a significant part of our vocabulary (and our documents) 
about them acknowledges the relevance of these, doesn't it? And with 
respect to the use of semantic relationships I think the UCR documents 
provides enough evidence of how important they are for the scope of SKOS...

Cheers,

Antoine

> Here is a draft response to Margie Hlava, comments welcome.
>
> --- begin draft message ---
>
> Dear Margie,
>
> Thank you for you support and your helpful comments. In response to
> your comment below:
>
> On Tue, Sep 30, 2008 at 12:00:44PM +0000, SWD Issue Tracker wrote:
>   
>> ISSUE-146: Last Call Comment: broadMatch and skos:narrowMatch should be used only when there are no exact or close matches for the term elsewhere?
>>
>> http://www.w3.org/2006/07/SWD/track/issues/146
>>
>> Raised by: Alistair Miles
>> On product: SKOS
>>
>> Raised by Margie Hlava in [1]:
>>
>> """
>> This is excellent work and generally a map can be made from a ANSI/NISO or 
>> a BSI or even and ISO thesaurus or controlled vocabulary standard to SKOS.  
>> However there are still a few confusions which prevent one from insuring 
>> complete interoperability from one to another.  This is represented by the 
>> section 10. Mapping Properties
>>
>> skos: broadMatch and skos:narrowMatch should be used only when there are
>> no exact or close matches for the term elsewhere?
>>     
>
> Questions of best practice such as this are deemed out of scope for
> the SKOS Reference. We hope that answers to questions such as this
> will emerge in the future within the community of practice, in
> response to implementation experience. We propose to make no change to
> the SKOS Reference, can you live with this?
>
>   
>> A taxonomic view of a thesaurus depends on the broader term narrower term 
>> relationships.  To SKOS this is not as important as the synonym (also known 
>> as equivalence) relationships.  The parent child, genus species, broader 
>> narrower term designations allows browse-able or navigational search.  This 
>> technique has taken much of the information industry  by storm for the last 
>> couple of years.  To allow this only as an after thought in SKOS is to 
>> marginalize an important area in findability.
>>     
>
> We believe the SKOS Reference makes no judgment as to the relative
> importance of the different types of relationship. Can you live with
> the document as-is?
>
> Kind regards,
>
> Alistair
> Sean
>
>   
>> [1] http://lists.w3.org/Archives/Public/public-swd-wg/2008Sep/0055.html
>>     
>
>   
Received on Wednesday, 22 October 2008 19:10:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 22 October 2008 19:10:14 GMT