Specializations of skos:broader in ISO25964

Hi,

This has become almost impossible for me to follow. The trick to use different fonts in the mail to know who said what and when doesn't work for me, clearly ;-)


>
> */[JDS3:>]ISSUE 6 – Modeling of BTG/NTG, BTP/NTP, BTI/NTI /*
>
> */We can introduce broaderGeneric, broaderPartitive and broaderInstance as sub-properties of skos:broader (see also [1]) /*
>
> Isn’t it better to call it broaderInstantive? (I think early SKOS drafts also used that name).


Good catch, Vladimir!
See http://www.w3.org/2006/07/SWD/track/issues/56


>
> */Further discussions are ongoing.  The proposal would be extended such that:/*
>
> */-  broaderGeneric is modeled as a sub-property of rdfs:subClassOf/*
>
> */- broaderInstance is modeled as a sub-property of rdf:type/*
>
> *//*
>
> Please don’t do this. It means you’ll equate skos:Concept with owl:Class, raise the level of required OWL inference, and engage Punning (because the same concept could be both a class and an individual).



Well, this will do it for *some* resources of type skos:Concept. It's not equating the class skos:Concept with the class owl:Class. There's a world apart...

(and yes, I believe that if someone uses broaderGeneric, that more-or-less says that that the involved concepts can be seen as classes, i.e. sets of resources. If this is not the case then probably their use of broaderGeneric doesn't comply with what the guidelines say)


>
> */ISO 25964 gives limits for the transitivity (and hierarchy) when using broaderPartitive/narrowerPartitive (whole – part) relationships./*
>
> */These should only be in a hierarchical relationship when it complies with the ALL/SOME test./*
>
> I agee with these principles (as illustrated by Jutta), but this is not fixable unless we give up skos:broaderTransitive.
>
> For our current purposes, this would be going too far (all TMS I know use skos:broader and all query expansions use skos:broaderTransitive),
>
> so for now let’s just define broaderGeneric, broaderPartitive and broaderInstantive as subprops of skos:broader, and leave it at that
>

I agree with the focus on the simple, non-transitive definition for BG, BP and BI.
But I think we should let the ISO WG (with the input from this community) the choice, whether they want to offer transitive super-properties (for BG and BP) if they think it can help data consumer. In the same way as SKOS introduced skos:broaderTransitive on top of skos:broader.

Antoine

Received on Monday, 11 November 2013 20:53:13 UTC