[Dspace-general] Google search results bypass metadata records

[Jodi Schneider]

On 6/25/07, Robert Tansley <roberttansley at google.com> wrote:
>
> The only real solution is to have backlinks in the PDFs etc -- the PDF
> you disseminate doesn't necessarily need to be a bit-perfect copy of
> your archival copy.  It wouldn't be too hard to build a Media Filter
> that copies the PDF and adds a link to the top of the copy; you could
> allow this PDF to be indexed rather than the archival copy.


I think this is worth discussing. Though I worry about separating the
dissemination copy from the archival copy, context (such as backlinks) is
important. Non-PDF bitstreams would also need to be considered.

It's a shame that there's no format like RTFD (RTF + directory bundle) which
would allow metadata to be distributed along with an archival bitstream.

I've been thinking about the "surround" context info wanted for a
dissemination copy. Here are my thoughts:
* basic item metadata (at least title, author, date) [because some items
don't self-disclose]
* link to the DSpace metadata/item page
* repository name, location, and link [because handles obscure provenance]
* collection name(s) and link [for awareness of related items]
* date of download (such as you see these days on downloads from nature.com)
[for citation purposes]

Using a MediaFilter would require time and/or space for generation/storage.
It might also decrease the dissemination of archival copies. Are there other
disadvantages?

-Jodi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/dspace-general/attachments/20070625/595125bb/attachment.htm