Re: Relationships involving collections

Hi Alasdair!

You wrote:

> With regard to the latest skos reference working draft, how should 
> mappings between vocabularies that involve collections be performed?
> 
> In the astronomy vocabularies that I have been working with, I have come 
> across several instances where I either need to directly relate 2 
> collections or a collection with a concept. One such example is for 
> relating the vocabulary of astronomy and astrophysics journal keywords 
> (A&A) [1] with the international astronomical union thesaurus (IAUT) 
> [2]. Below are brief snippets of the two vocabularies.
> 
> A&A
> Concept: "Sources as function of wavelength"
>  NT Collection: "Gamma Rays"
>    NT Concept: "Gamma ray bursts"
>       Concept: "Gamma ray observations"
>       Concept: "Gamma ray theory"

That's in RDF:

aa:1 a skos:Concept ;
  skos:prefLabel "Sources as function of wavelength" ;
  skos:narrower aa:2 .

aa:2 a skos:Collection ; rdfs:label "Gamma Rays" ;
  skos:member aa:3 ;
  skos:member aa:4 ;
  skos:member aa:5 .

aa:3 a skos:Concept ; skos:prefLabel "Gamma ray bursts" .
aa:4 a skos:Concept ; skos:prefLabel "Gamma ray observations" .
aa:5 a skos:Concept ; skos:prefLabel "Gamma ray theory" .

> IAUT
> Concept: "Radiation"
>  NT Concept: "Gamma rays"

That's in RDF:

iaut:1 a skos:Concept ; skos:prefLabel "Radiation" ;
   skos:narrower iaut:2 .
iaut:2 a skos:Concept ; skos:prefLabel "Gamma rays" .

> I would like to assert
> A&A:"Gamma Rays" skos:exactMatch IAUT:"Gamma rays"

You try to assert

aa:2 skos:exactMatch iaut:2 .

But why don't you assert the following:

iaut:2
   skos:narrowMatch aa:3 ;
   skos:narrowMatch aa:4 ;
   skos:narrowMatch aa:5 .

> In fact, as I have typed up this example I wonder if the A&A vocabulary 
> snippet I have given is in fact valid in the new skos reference as is 
> declares a collection to be a narrower term and this goes against the 
> domain and range declarations for the BT/NT relationships.

Yes, mapping a collection and a concept is not valid - but in practise 
(as you have shown) it would be very useful. We could either broaden the 
  domain and range of the mapping vocabulary or add a second mapping 
vocabulary. In both cases you MUST declare some rules:

<A> a skos:Concept .
<B> a skos:Concept .
<C> a skos:Concept .
<X> a skos:Collection .

<A> skos:narrower <X> .
  <X> skos:member <B> .

<X> skos:exactMatch <C> .

entails

<C> skos:narrowMatch <B>


This could get a bit more complicated (for instance transitive member 
relations) but makes sense.

Thanks for pointing to this issue.

Greetings
Jakob

-- 
Jakob Voß <jakob.voss@gbv.de>, skype: nichtich
Verbundzentrale des GBV (VZG) / Common Library Network
Platz der Goettinger Sieben 1, 37073 Göttingen, Germany
+49 (0)551 39-10242, http://www.gbv.de

Received on Tuesday, 5 February 2008 12:11:12 UTC