Tuesday, October 9, 2012

New Bibliographic Framework: Update with Eric Miller

Posting to Autocat

I just listened to Eric Miller's talk at the Library of Congress about the new bibliographic framework. http://www.loc.gov/today/cyberlc/feature_wdesc.php?rec=5605 A lot of it was quite interesting and I agreed with most of it. Eric Miller (formerly of W3C Semantic Web group, formerly with OCLC, now with a company he founded, Zepheira) has a special interest in linked data and of course, that is on everyone's mind with the new bibliographic framework.

He started with the normal, fantastic, utopian possibilities for linked data that we have gotten used to seeing now, and after that, it got interesting. At least as I understand his idea for library linked data, he sees it primarily as a way to hook together databases "similar" to ours, such as museums, archives--I would imagine open archives too--and a few others. He does not believe that linked data will bring libraries and the business world into such a close collaboration. The goal of libraries working closely with archives and museums seems to me to be very positive and achievable, but rather limited and probably could have been done a decade ago or even more, even without linked data. Still, he may be absolutely right about staying more or less separate from the business world.

In my opinion however, even if this is done, the public will still be stuck with the same online versions of card catalogs that are incomprehensible to them: the subject headings will still make no sense and the cross-references for authorized forms of names still won't work. Nobody wants to tackle these fundamental problems of the catalog yet, so I see relatively little actually changing.

As I have maintained for a long time, going toward linked data is fine and there is nothing wrong with it--but we shouldn't expect too much from it. It is not a solution for anything at all. The actual catalog will still be based on browsing the card catalog and will still be obsolete. With such a limited group as Eric Miller proposes, there may be a chance for the whole to adhere to some kind of bibliographic standards.... maybe.

He was asked about RDA and he said that it is not needed to enter the world of linked data and that several agencies have achieved linked data without it. Also, that nobody outside the library community has any idea about work, expression, manifestation or item. One very positive point he made was that the new bibliographic format to enter the semantic web/linked data world *must be simple*. Because if the format is too complex, nobody will want to implement it. Absolutely true. Of course, such simplicity may be OK to get libraries into the linked data universe, but it will not be adequate for the libraries themselves who will always need much more. As I said, interesting.

Yet, maybe this could mean that we could get into linked data as soon as possible so that we can then begin to move forward.

I suggest this talk for everyone.

No comments:

Post a Comment