On Mon, 6 Sep 2004, Alistair Miles wrote: >dc:identifier has recently occurred to me here also. > >What if we go ahead with the deprecation of skos:externalID, and replace with >dc:identifier? (And drop the proposal for a new property skos:localID). > >The problem of how to cope with concepts that may have more than one identifier >(possibly deriving from different schemes) still remains, however. I don't see the problem. There is no reason to restrict the cardinality (number of dc:identifiers), so you could do as you suggested, or as I have modified it below: >What if we recommend here that a sub-property of dc:identifier is created by the >scheme owners for each identifier? So e.g. (I've modified Al's example -CMN) <skos:Concept rdf:about="http://example.org/topics/098"> <ex:conceptNumber>098</ex:conceptNumber> <ex2:id>A.34.12</ex2:id> <dc:identifier rdf:parsetype="resource"> <dc:scheme rdf:resource="http://purl.org/dc/elements/1.1/"/> <rdf:value xml:lang="en">Title</rdf:value> </dc:identifier> <dc:identifier xml:lang="fr" rdf:value="Nom du truc"/> </skos:Concept> (I should check the syntax, but the idea is the important thing :-) >Where ex:conceptNumber and ex2:id are both hypothetical properties created by >scheme owners, and both are declared to be subPropertyOf dc:identifier by scheme >owners. (Scheme owners can also declare these props as inversefunctional if >they choose). >Comments? Made sense to me... ChaalsReceived on Tuesday, 7 September 2004 08:22:20 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:38:52 GMT