[Proposal][SKOS-Core] Collections

I'm thinking, we really ought to allow people say things like (with proposed
terms under the 'proposed' namespace prefix) ...

  <skos:Concept rdf:about="C1">
    <skos:prefLabel>Aircraft</skos:prefLabel>
    <skos:narrower>
      <proposed:Collection rdf:about="colA">
        <rdfs:label>Aircraft by function</rdfs:label>
        <proposed:members rdf:parseType="Collection">
          <skos:Concept rdf:about="C10"/>
          <skos:Concept rdf:about="C11"/>
          <skos:Concept rdf:about="C12"/>
        </proposed:members>
        <proposed:ordered>true</proposed:ordered>
      </proposed:Collection>
    </skos:narrower>
  </skos:Concept>

... and with the rdfs:label on a proposed:Collection being entirely
optional.

I.e. the above example could just as easily have had 'skos:broader' or
'skos:related' instead of 'skos:narrower'.

This suggestion does have some major implications:

(1) It breaks the current range/domain restraints on sub-properties of
skos:semanticRelation (such as skos:narrower) if we allow them to be used
with a Collection as the object of a statement.

(2) We want to infer from the above example that ...

	'For all m such that m is a member of collection c:
	(x,p,c) -> (x,p,m)'

   ... the problem is that we can't write that in a rule, because of the way
rdf collections are structured.  

We could get around that, by forcing people to make statements e.g. 
	
  <skos:Concept rdf:about="C12">
    <proposed:isMemberOf rdf:resource="colA"/>
  </skos:Concept>

which would allow a rule to be written:

(x,p,c)(m,proposed:isMemberOf,c) -> (x,p,m)

  ... but the point is, we would be relying on doing some inference to
obtain the pure network of semantic relations between concepts, and a tricky
sort of inference at that.

Anyway, food for thought.

Al.



---
Alistair Miles
Research Associate
CCLRC - Rutherford Appleton Laboratory
Building R1 Room 1.60
Fermi Avenue
Chilton
Didcot
Oxfordshire OX11 0QX
United Kingdom
Email:        a.j.miles@rl.ac.uk
Tel: +44 (0)1235 445440

Received on Tuesday, 24 August 2004 14:28:36 UTC