At 12:31 PM 7/26/99 +0200, Bernhard Eversberg wrote:
>
>We do not know, at present, what kinds and degrees of inconsistencies can
>be tolerated in databases without degrading their quality for discovery
>too much.
We do have plenty of experience with imperfect metadata, although I agree
that we've never quantified the effects of that imperfection. I've just
completed a study of cross-catalog searching -- catalogs with full MARC
records and AACR2 rules -- that got results that bordered on random. The
factors that affected results were many: differences in metadata,
differences in index choices, differences in search methods. I agree that
you have more of a fighting chance if your metadata is good quality and
consistent, but there are so many factors that go into resource discovery
that we can't put all of our eggs into the metadata basket. (How's that for
Weibel-izing a sentence?) At some point, we need to quit fine-tuning the
metadata and get on to the act of discovery, which has some interesting
problems of its own.
----------------------------------------------
Karen Coyle [log in to unmask]
University of California Digital Library
http://www.kcoyle.net 510/987-0567
----------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|