Casey Bisson wrote:
|Diane I. Hillmann wrote:
|> Why can't we just use OAI to pass them around? Much less overhead,
|> and the aggregators and services can develop on their own once the
|> records are available. ...
|
|A problem here is license. While many of us are creating some records
|and fixing many more, our catalogs are filled with records derived
|from any number of sources and controlled by a greater number of
|licenses. We can't legally share our records (especially into a copy-
|left collection) if we don't own them/control the license.
The thing to do would be to get a significant number of large libraries,
both public and academic, to cull out of their local databases all the
bib records where they are the originating library (tag 040 subfield a).
(Since they are the originating library, their local record would not
have any possibly questionable shared-database additions or
modifications from a copyright standpoint but would almost certainly be
"pristine"--unless they redownloaded for some reason.) The LC bib
database could be included, too, since it's freely available to
Americans--but I believe there are restrictions to allowing access to
non-U.S. libraries, since it was created with American tax money and not
funds from other countries. (You'd have to check on how to deal with
that.) Anyway, that combination ought to give a good start to such a
cooperative project. Of course, other considerations come into play
beyond what have already be mentioned: if two or more libraries (in
different shared cataloging cooperatives) created bib records from
scratch for the same item, which record goes into the repository?
What's the process for correcting errors, and who has authority to do
so? Or is the repository like a wiki, where anybody can change records
willy-nilly? How would duplicate records, if any, be handled? Just
read any shared cataloging cooperative's manual(s) to get an idea of all
that's involved. It's easy to look at big pictures, but (in a digital
world) little pixels all have to fit together correctly for the big
picture to exist. I'm rather sure the wished-for repository is quite
possible, but it won't be large unless *lots* of libraries (especially
big ones) participate. For example, the current typical rate for local
creation (from scratch) of book bib records in OCLC is about 2% (AV is
around 10%-20%)--at least in a public library environment. That's not a
lot of books that any one typical local institution can contribute (for
us, that would be only around 7000-7500 titles). Compare that to the
4-5 million or more records (I no longer know how many LC MARC records
exist) created by the Library of Congress since 1968. Without the LC
bib database, you're likely talking a piddly-sized repository--nothing
approaching OCLC's 70+ million bibs without all the other article
records using up numbers. FWIW.
Harvey
--
===========================================
Harvey E. Hahn, Manager, Technical Services Department
Arlington Heights (Illinois) Memorial Library
847/506-2644 - FX: 847/506-2650 - Email: hhahn(at)ahml(dot)info
OML & Scripts web pages: http://www.ahml.info/oml/
Personal web pages: http://users.anet.com/~packrat
Received on Wed Apr 25 2007 - 17:00:55 EDT