> I wonder if anyone has any thoughts or suggestions on how to treat
> digitised items such as short stories, reports, etc. Presuming that most
> of the information will be the same from page to page in a short story, is
> it possible to have one DC record for the whole story( multiple digital
> files) and indicate the various pages as " Has Parts" or is it necessary
> to have a DC record for each page or digital file and indicate the
> relationship with the previous page as "Is Part of".
It depends on how the DC info is being associated with the files. In the
case of RDF all resources are identified by URIs and as such a URI can be
associated with either the article as a whole, or an individual part. If the
RDF is in an external file then one RDF file can serve to describe all of
the relevant article parts. In the case of embedded information you may wish
to replicate at least the "Is Part Of" to assist discovery of the entire
information set.
There is a problem with the common scenario where a URI can be seen as
identifying both the article as a whole, and the first part. In some such
cases there is a redundant URI identifying the first part (e.g. the urls are
something like .../art.asp?art=34, .../art.asp?art=34&part=2,
.../art.asp?art=34&part=3, where .../art.asp?art=34&part=1 could be used,
but generally isn't), and then you could use the redundant URI to identify
part 1, and the more concise URI to identify the entire article.
That's only one way of dealing with one kind of article <-> URI mapping with
one way of distributing DC information, but it may be a start towards
something suitable for your uses.
|