Re: New Draft comments: Motivations in SKOS

Hi Rob,

This looks good!
The only issues are a small typo (AnotherSchemne->AnotherScheme), the fact that I would have lower-cased the instance ids (with my RDF eyes, my first reaction would be to read oa:MotivationSheme as a subclass of skos:ConceptScheme, not an instance of it ;-) ).

Also, I think all the broadMatch can be replaced by broader: the semantic relations are embedded in the design of the concerned concepts, they are not post-ante reconciliation of concepts that were created in isolation.

On comment 1: I agree for keeping oa:Motivation makes much sense. But part of my point was to get rid of the general oa:annotating concept. Asserting that a concept is narrower than oa:annotation doesn't had much information to asserting that this concept is a member of oa:motivationScheme, I think.

Cheers,

Antoine



>
> Hi Antoine, and all,
>
> Could you verify that the below RDF is what is intended by your suggestions?
>
> -------------------------
> oa:MotivationScheme a skos:ConceptScheme ;
> skos:hasTopConcept oa:annotating.
>
> oa:Motivation rdfs:subClassOf skos:Concept .
>
> oa:annotating a oa:Motivation ;
> skos:prefLabel "Annotating"@en .
>
> oa:editing a oa:Motivation ;
> skos:inScheme oa:MotivationScheme ;
> skos:broadMatch oa:annotating ;
> skos:prefLabel "Editing"@en .
>
> new:correcting a oa:Motivation ;
> skos:inScheme new:AScheme ;
> skos:broadMatch oa:editing ;
> skos:prefLabel "Correcting a Mistake"@en .
>
> new2:fixing a oa:Motivation ;
> skos:inScheme new2:AnotherSchemne ;
> skos:broadMatch oa:editing ;
> skos:closeMatch new:correcting ;
> skos:prefLabel "Fixing a Mistake"@en .
> ---------------------
>
> And to your comments...
>
> On Sun, Jan 6, 2013 at 8:59 AM, Antoine Isaac <aisaac@few.vu.nl <mailto:aisaac@few.vu.nl>> wrote:
>
>     1. Is oa:Annotating really needed? It should be enough that motivations are in oa:MotivationScheme (or are defined to be a sub-concept of a concept that is in oa:MotivationScheme) to infer that these are kind of annotation purposes.
>
>
> The thought was that it would be easier to assert that the range of oa:motivatedBy is a oa:Motivation, rather than any skos:Concept. Also, as below, they may not be in any Concept Scheme. Unless there's a reason not to, I think this would be good to keep.
>
>     2. Using skos:topConceptOf is valid, but this property was coined for technical reasons. It would be better to keep to the property in the other direction, skos:hasTopConcept.
>
>
> Fixed in to-be-published revised draft.
>
>
>     3. I am not sure that putting all new motivation concepts (new:Correcting and new2:Fixing) in the reference oa:MotivationScheme. If several applications create their own, potentially overlapping motivation concepts, then oa:MotivationSheme risks becoming difficult to use. For extensions, knowing that a concept defined to be a sub-concept of a "reference" concept that is in oa:MotivationScheme (possibly indirectly, via skos:broaderTransitive) should meet most requirements I can think of
>
>
> Agreed, fixed this.
>
>     4. It is good practice to use language tags with SKOS labels. This should appear in the machine-readable file, but also be reflected in the example.
>
>
> Fixed.
>
>     With all these suggestions, Figure B could be reworked to look more like the attached diagram (not trying to enforce any graphic convention here! It's just that I don't have time to refine it...)
>
>
> This, too, will be fixed and published very shortly :)
>
> Rob

Received on Thursday, 10 January 2013 08:45:18 UTC