> From [log in to unmask] Fri Apr 26 07:39 MET 2002
> Mime-Version: 1.0
> Content-Disposition: inline
> Date: Fri, 26 Apr 2002 17:38:45 +1200
> From: Douglas Campbell <[log in to unmask]>
> Subject: DCQ in RDF/XML clarifications
> To: [log in to unmask]
> Content-Transfer-Encoding: 8bit
> X-MIME-Autoconverted: from quoted-printable to 8bit by scarlett.mathematik.Uni-Osnabrueck.DE id HAA00404
>
> Kia ora,
>
> I'm confused about a couple of things in the "Expressing Qualified Dublin Core in RDF / XML" proposal [1]. I looked through the draft DC Terms RDF Schema in the appendix, but not knowing enough about RDF I couldn't figure them out.
Such things might become addressed in the planned "gentle intro" - some points are given below...
> I had a look thru "Dublin Core in RDF Draft" [2] and "RDF Model and Syntax Specification" [3] but unfortunately the answers didn't jump out at me... :-(
>
> 1. Use of rdf:value elements for encoding schemes
>
> In the draft DC Terms RDF Schema, the comments for the five Subject encoding schemes state an rdf:value MUST be provided, but none of the other encoding schemes for other DC15 elements mention this. Am I correct in assuming it is optional for the other DC elements?
The question is addressed in 2.3.4 - one should provide at most one.
>
> 2. How DCMIType data should appear in RDF
>
> At first I assumed DCMIType encoded data would look like any other encoding scheme, eg:
>
> <rdf:Description about="...">
> ...
> <dc:type>
> <dcterms:DCMIType>Image</dcterms:DCMIType>
> </dc:type>
> ...
> </rdf:Description>
>
> But on closer inspection I noticed the draft DC Terms RDF Schema description of Type is different to the other encoding schemes, and I couldn't quite figure out what is intended - unfortunately the example in 2.3.12.2 doesn't look like a "typical" DC record to me.
see 4. below
>
> So I'm wondering which (if any) of the following it should be:
>
> 1.
> <dc:type>
> <dctype:Image/>
> </dc:type>
>
> 2.
> <dctype:Image/>
> [ie. you don't need to put the dc:type in as it is implied by using "dctype:Image"]
>
> 3.
> <rdf:Description about="...">
> <dctype:Image>
> ... [the rest of the DC record]
> </dctype:Image>
> </rdf:Description>
> [ie. you wrap the DC record with the type, kind of like the example in 2.3.12.2]
>
4. <?xml version="1.0">
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:dctype="http://purl.org/dc/dcmitype/"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<dctype:Image rdf:about="urn:abc:1234567">
<dc:creator>Carl</dc:creator>
<dc:description>A picture of foo</dc:description>
</dctype:Image>
</rdf:RDF>
The case of dcmitype is indeed special:
It's (currently) the only case of a dcmi namespace other than
/elements/1.1/ and /terms/ -
More important: The RDF Schema for this namespace declares the entries
of /dcmitype/ as rdfs:Class(es) AND the properties rdf:type and dc:type
are (closely) related.
This makes it's entries available as RDF typed nodes and allows for
a rather condensed presentation in RDF. In particular you
can get rid off the pedestrian rdf:Description envelope:
<?xml version="1.0">
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<rdf:Description rdf:about="urn:abc:1234567">
<rdf:type rdf:resource="http://purl.org/dc/dcmitype/Image"/>
<dc:creator>Carl</dc:creator>
<dc:description>A picture of foo</dc:description>
</rdf:Description>
generates the graph described in 4. also.
There is a loss of information which comes by
http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/#usage
in particular C.2. Namespace URIs:
"...In addition, the namespace name serves as the identifier for the corresponding
RDF Schema. ...An RDF processor can expect to use the schema URI to
access the schema content...."
Accessing the schema supplies you with a label: "Image",
which you might want to make use of and the human understandable
semantics of the resource denoted by
http://purl.org/dc/dcmitype/Image
You can water (dumb) down the information to
<rdf:Description rdf:about="urn:abc:1234567">
<dc:type rdf:resource="http://purl.org/dc/dcmitype/Image"/>
<dc:creator...
<dc:description...
</rdf:Description>
Now you lost, that "http://purl.org/dc/dcmitype/Image" has
anything to do with RDF.
This version is (almost) off qualified DC: all
what is left is, that you still know the character string
http://purl.org/dc/dcmitype/Image is a URI -
Still slightly above ground...
Best,
rs
> Thanx,
> Douglas Campbell
> National Library of New Zealand
>
> [1] http://dublincore.org/documents/2001/11/30/dcq-rdf-xml/
> [2] http://logicerror.com/dcrdfDraft
> [3] http://www.w3.org/TR/REC-rdf-syntax/
>
|