Is this the use case for RDFa?

[Patrick Durusau <patrick@durusau.net>]

Greetings!

Still reviewing all the posts in preparation for today's meeting and 
may, emphasis on may, have a little better understanding of why Bruce 
and Elias are advocating an RDFa (although specified in ODF) approach.

Consider that I have a subject, one of John's patients and what follows 
in the document are a number of statements about that patient.

If I use the xml:id approach to mark the patient, in this case, say 
<span xml:id="snowfall_1">Patrick Durusau</span> and then later in the 
file there are other statements about this patient, omitted here for 
medical privacy reasons, ;-), the question is how do those properties 
"roll up" to be associated with that subject?

Is that an accurate statement?

But, if I use an rdf:about attribute on a container, then those later 
statements, according to RDFa will "roll up" to that subject?

Hmmm, that sounds to me like a question of how the properties that 
represent the object and predicate of the triple are to be associated 
with the property that identifies the subject.

And the argument that I understand Elais and Bruce to be making is that 
the "roll up" is easy enough to see with the RDFa approach, whereas 
using xml:id requires additional mechanisms. But both are representing 
the same information.

Interesting.

Which sheds some light on Elias's position on non-duplication of 
content. It is "re-use" of the content inline so as to not replicate it 
in the metadata file. But I suppose nothing prevents an application that 
mandates external storage only, from transforming the "inline" metadata 
into metadata in one or more metadata files.

Thinking of the use case where metadata is not shared on the transfer of 
the document.

Oh, and thinking about the enhanced search case, note that for the 
general case, declaring a vocabulary for a document, all we need to be 
able to do is point from the metadata file to a declaration of that 
vocabulary. Doesn't really matter where or how it is held, although I do 
think we need to specify at least one encoding for such files.

The closer case to being relevant here is how do we point out if a term 
uses a particular vocabulary or has a meaning that is not consistent 
with the vocabulary declared for the document? Obviously the xml:id 
mechanism isn't going to be much help here, unless we say that the 
pointing to a vocabulary is done in the metadata file and that any 
specialized term is pointed to from that file via the xml:id mechanism.

Hope this helps clarify and doesn't confuse the issues!

Hope everyone is having a great day!

Patrick

-- 
Patrick Durusau
Patrick@Durusau.net
Chair, V1 - Text Processing: Office and Publishing Systems Interface
Co-Editor, ISO 13250, Topic Maps -- Reference Model
Member, Text Encoding Initiative Board of Directors, 2003-2005

Topic Maps: Human, not artificial, intelligence at work!