RE: RDF Thesauri and KOS API

Hi Alistair

First thank you for acting on my comments about human readability so
swiftly.

> > In addition, the other thing which I would help my 
> understanding is a
> > description of how you expect these thesauri to be processed? 
> 
> These kinds of questions I think fall under the remit of our 
> development of a web API for interacting with a thesaurus 
> service.  My preference is to keep such an API wholly 
> concept-oriented.  So the interaction might look
> like ...
> 
> User enters search string (regular expression)
> -> Thesaurus server returns all concepts with labels that 
> match the query.
> User requests broader/narrower/all relatives of a specific concept
> -> Thesaurus server returns requested concepts
> 
> Such an API could easily include the option to return a closure of all
> narrower concepts, or just the direct narrower concepts.  
> This could be a
> parameter of the method call cf. Jena API.
>  
> Does this answer your question?

Well I guess the problem is 
http://www.w3c.rl.ac.uk/SWAD/deliverables/8.1.html
isn't complete, there are still placeholders for examples, so the problem is
I can't fully understand the schema without those examples. So to take your
scenario

> User requests broader/narrower/all relatives of a specific concept
> -> Thesaurus server returns requested concepts

it's not clear to me what constitutes the "requested concepts" that are
returned here, for example

- are the broader terms transitive? ditto for narrower and related?

- you say that that broaderGeneric is equivalent to rdfs:subClassOf - does
this mean that this variant of broader is transitive, but the others aren't?

- can you illustrate the differences between broaderInstantive,
broaderGeneric and broaderPartitive? ditto for the narrower variants?

Perhaps if you add the examples and briefly discuss them in the text then
this may be sufficient to clarify these issues?
Alternatively, another way to clarify things would be to have an example
thesauri that artificially uses all the constructs you define to illustrate
your schema?

Hope this helps, kind regards

Dr Mark H. Butler
Research Scientist                HP Labs Bristol
mark-h_butler@hp.com
Internet: http://www-uk.hpl.hp.com/people/marbut/

Received on Thursday, 27 November 2003 05:55:23 UTC