- From: Miles, AJ \(Alistair\) <A.J.Miles@rl.ac.uk>
- Date: Tue, 11 Oct 2005 12:43:03 +0100
- To: <DC-RDF-TASKFORCE@jiscmail.ac.uk>
- Cc: <public-esw-thes@w3.org>
Hi Andy, > We had the need for using other subject encoding schemes with > Dublin Core, other than the those listed by DCMI, e.g., DDC, > LCSH, MESH, UDC. In our discussion with Andy Powell and Stu > Weibel they indicated that DCMI was not happy with subject > encoding schemes since everytime you need a new one, you have > to add it to the Dublin Core specification. Last year we > proposed an alternate, backward compatible, method for DCMI > to deprecate the use of subject encoding schemes and allow > any subject scheme to be specified via URI. I'm not sure > where or how far the proposal went. Yes, I think we need to work out how 'subject schemes' should be declared in principle, allowing DCMI or anyone else to declare and use their own within DC metadata. > > Our proposal produced the same RDF graph structure as the > following example, lifted from the DCMI document: > > <rdf:RDF > xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" > xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" > xmlns:dc="http://purl.org/dc/elements/1.1/" > xmlns:dcterms="http://purl.org/dc/terms/" > > > <rdf:Description> > <dc:subject> > <dcterms:MESH> > <rdf:value>D08.586.682.075.400</rdf:value> > <rdfs:label>Formate Dehydrogenase</rdfs:label> > </dcterms:MESH> > </dc:subject> > </rdf:Description> > </rdf:RDF> > > Basically, the above produces the following raw RDF: > > <rdf:RDF > xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" > xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" > xmlns:dc="http://purl.org/dc/elements/1.1/" > xmlns:dcterms="http://purl.org/dc/terms/" > > > <rdf:Description> > <dc:subject> > <rdf:Description> > <rdf:type>http://purl.org/dc/terms/MESH</rdf:type> > <rdf:value>D08.586.682.075.400</rdf:value> > <rdfs:label>Formate Dehydrogenase</rdfs:label> > </rdf:Description> > </dc:subject> > </rdf:Description> > </rdf:RDF> > Nit-pick: the rdf:type element should be... <rdf:type rdf:resource="http://purl.org/dc/terms/MESH"/> ... see [1]. > Thus, you can deprecate the use of use of subject encoding > schemes and allow new subject encoding schemes without having > to add them to Dublin Core by changing the URL in rdf:type, > for example GSAFD (a list of genre terms for drama and fiction): > > <rdf:RDF > xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" > xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" > xmlns:dc="http://purl.org/dc/elements/1.1/" > xmlns:dcterms="http://purl.org/dc/terms/" > > > <rdf:Description> > <dc:subject> > <rdf:Description> > <rdf:type>info:kos/scheme/gsafd/20010920</rdf:type> > <rdf:value>GSAFD000001</rdf:value> > <rdfs:label>Adventure fiction</rdfs:label> > </rdf:Description> > </dc:subject> > </rdf:Description> > </rdf:RDF> > My guts tell me it should either be ... OPTION 1: <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:dc="http://purl.org/dc/elements/1.1/"> <rdf:Description rdf:about="http://www.example.com/somedoc"> <dc:subject rdf:datatype="http://purl.org/dc/terms/MESH">D08.586.682.075.400</dc:subject> </rdf:Description> </rdf:RDF> ... or OPTION 2: <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:dc="http://purl.org/dc/elements/1.1/"> <rdf:Description rdf:about="http://www.example.com/somedoc"> <dc:subject rdf:resource="http://purl.org/dc/terms/MESH/D08.586.682.075.400"> </rdf:Description> </rdf:RDF> ... and (stated outside of 'instance metadata') ... <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:dc="http://purl.org/dc/elements/1.1/"> <rdf:Description rdf:about="http://purl.org/dc/terms/MESH/D08.586.682.075.400"> <rdfs:label>Formate Dehydrogenases</rdfs:label> <rdf:value>D08.586.682.075.400</rdf:value> <rdfs:isDefinedBy rdf:resource="http://purl.org/dc/terms/MESH"/> </rdf:Description> </rdf:RDF> Option 1 requires little or no modification to the AM, but does require a description of how to map from datatyped literal values to SKOS-like resource descriptions. Option 2 requires some modification to the AM, but does not require any additional mapping to SKOS-like resource descriptions, because it already is a SKOS-like resource description, and if you want to use SKOS you only have to add statements. > Not sure whether this helps your situation or not... > > > > > Also, the pattern: > > > > > > <http://www.example.com/somedoc> dc:subject > > 'D08.586.682.075.400'^^dcterms:MESH. > > > > > > ... seems to also match closely the XML encoding of qualified DC, > > using xsi:type. > > > > The use of xsi:type for the XML encoding of qualified DC > seems, to me, to be a hack and doesn't quite align with its > intended purpose in XML schema. Also, some XML parsers choke > on anything in xsi:type that isn't a qualified XML schema type. I was wondering about that. Cheers, Al. [1] http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/#section-Syntax-typed-nodes
Received on Tuesday, 11 October 2005 11:43:18 UTC