- 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