just to make things tricky, I want to remind you of the needs of the
accessibility community.
We tend to think of resources as composites of objects, some of which
may be inaccessible. In this case, we substitute an alternative or
sometimes we augment the original with an extra object. Now we don't
think of these things as different versions, as such, but as
alternatives or equivalent alternatives - the latter when, for example,
I use something completely different from the original for a blind
person because even a description of the image would not really help.
So, we are talking about primary and equivalent objects .....these are
not fixed labels so that what is a primary in one case may be an
equivalent for a different primary. Also, both the primary and the
equivalent are sometimes embedded within the same object ...
hope this adds to the fun!
Liddy
On 13/01/2005, at 4:17 AM, Diane Hillmann wrote:
> Folks:
>
> I'm involved in an effort to define "Best Practices" for OAI-PMH
> interactions, sponsored by the Digital Library Federation and
> including some of the most experienced practitioners in the business.
> Despite all that collective knowledge, we're currently enmeshed in a
> discussion on how to advise data providers to deal with versions,
> given the Dublin Core 1:1 rule. I'm sure the fact that we are not in
> agreement will come as no surprise to anyone.
>
> It seems to us that one major difficulty is that, for those who
> choose to comply with 1:1 in the spirit of the rule, if not in the
> letter, each provider will choose to slice and dice their content
> differently, present the metadata for it differently, and assume that
> their "best solution" works for downstream users. These solutions tend
> to be very creative, but in fact they often create considerable
> problems for indexing and access for aggregators of metadata.
>
> Although the OAI-PMH minimum requirement for metadata is Simple DC,
> we've decided to focus our quest on solutions for Qualified DC first,
> and then consider how "dumb down" would work in that context. So,
> we're asking the folks on this list for discussion and examples of how
> they deal with versions, particularly versions of the same (or what
> they consider the same) intellectual object. We'd like to be able to
> include some of these solutions in our practices documents, with some
> analysis of what might be the advantages and disadvantages from the
> point of view of the provider and downstream user.
>
> Thinking longer term, I was struck recently by an interchange on the
> DC Architecture list about "description sets" in relation to the DC
> Abstract Model (DCAM in these postings).
>
> http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0501&L=dc-
> architecture&T=0&F=&S=&P=60
> http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0501&L=dc-
> architecture&T=0&F=&S=&P=288
> http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0501&L=dc-
> architecture&T=0&F=&S=&P=401
>
> It seemed to me that there may be something in the concept of
> "description sets" that we could consider in this context, and, at the
> very least, propose a best practice that conforms in essence to the
> approach taken in the abstract model. I'd be interested in hearing
> any comments on this possibility.
>
> Regards,
> Diane
>
>
>
>
> -*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-
> *-*-*-*-*-*-*-*-*-*-*-*-*-*
> Diane I. Hillmann
> Director of Library Services and Operations
> National Science Digital Library
> Cornell Information Science Voice: 607/255-5691
> 301 College Ave. Fax: 607/255-5196
> Ithaca, NY 14850 Email: [log in to unmask]
> -*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-
> *-*-*-*-*-*-*-*-*-*-*-*-*-*
|