Neil Godfrey wrote:
>
> 2008/6/25 Ian Stuart <[log in to unmask] <mailto:[log in to unmask]>>:
>
>
> Having spent some time developing a base-line repository, and a fair
> amount of time thinking about how repositories interact with both
> users and each other, I (personally) can't see how SWAP (with DC,
> RDA, or RDF) is going to work **given the current technologies**
>
>
> What is the current state of play with SWAP and the modules developed
> for Eprints, DSpace and Fedora repositories, exactly?
I can't answer specifically for DSpace or Fedora, as I don't use their
systems, however pretty much every repository works the same way:
An item is deposited, and the user fills in various fields of data. This
is stored as rows in (one or more) tables in a database.
In EPrints, various columns in the table correlate to fields in the SWAP
structure, thus it "should" be possible to do a complex data extract and
construct a representation of all Items for a given Work.
I say "should", as this relies on matching text fields... not the most
definitive, or accurate, method for determining equality.
In my view, unless you build the hierarchical SWAP relationships from
the outset, you will never actually be able to build a proper SWAP tree
(and there is a problem with swap too: in research, an academic has a
topic of research they are interested in. To investigate that topic,
they apply for grants to research various aspects of that topic. Thus
multiple "Works" relate to a common "Topic", and are rarely isolated
items on their own...)
--
Ian Stuart.
Bibliographics and Multimedia Service Delivery team,
EDINA,
The University of Edinburgh.
http://edina.ac.uk/
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|