On 11/22/05, Barbara B Tillett <[log in to unmask]> wrote:
> Bruce - One of the trends in future cataloging codes is to take metadata from whatever source, as is, and to build on that to meet the FRBR identify, select, and obtain tasks, with controlled access points and uncontrolled keywords for access to meet the 'find' task....
Good; that makes sense. RDF would be *perfectly* suited to these goals in fact.
> Taking "citation" information would be part of that grabbing of descriptive metadata for the manifestation, so if you're working on automatic generation of the citation, isn't that part of the same thing?
Yes. One of things I talked about at Access 2005 -- my take home
point in fact -- was that the recent discussion around metadata in
OpenDocument could allow us to complete the metadata cycle. We remove
the disconnect between authoring, finding and consuming metadata.
E.g. let's say in the future I'm authoring a paper wth an
OD-compatible application. When I add citations to the document, the
RDF metadata description is embedded in the file archive. When I send
it to a publisher, they can extract that metadata, and embed it in the
PDF files they produce (Adobe is using an RDF subset in their
applications, and so one can add the same metadata to differnent file
formats; XML, PDF, TIFF, etc.). You then have "smart documents" that
tools can then use to provide better functionality to user. E.g. I
should be able to copy-and-paste content from such a PDF into some
document somewhere, and the metadata travels with it.
Obviously, library metadata could play a big role in this sort of vision.
But practically speaking, I do think you need RDF to enable this (the
seamless merging of metdata). XML alone simply won't work.
So I think if anyone actually wants to look into this, you need to
start over and:
1) get rid of XML Schema. It is not suited technically to doing this
sort of work.
2) spend some time looking at best practice RDF stuff like SKOS, FRBR
(which is still incomplete, but in development) and FOAF
3) maybe engage with people like Ian Davis (one of the authors of the
FRBR in RDF schema, who now works as technical lead for Talis) to find
a way forward on all this
> There is another aspect of 'citation' in future cataloging rules, that is, naming the works/expressions in a unique text string (now a uniform title, but hopefully in a compatible but better structure in the future). - Barbara
This would be REALLY useful, in fact, and is a big issue! Citations
are tricky, though, because they're often ambiguous about the FRBR
level. E.g. if I say (Doe, 1999, p34), that page number is already
pointing at the manifestation level.