Till Kinstler wrote:
>
> Linked data ist not about exposing library catalogues to google on some
> stupid way. Its about making this data "identifiable" (and
> "addressable") for any kind of use (typically OPACs totally fail on
> that, there is no simple way to address a record ...
or a set of records! That's to say, to extract a result set from a
catalog for software and then do something with. At least, catalogs
could provide standardized links by which a user can jump in and see the
results. WorldCat does that, Pica too, LC not. WorldCat's results can
of course not be processed by software for any meaningful results.
OTOH, Z39.50 was done with the intention to develop just that kind of
capability! But in pre-Web times. OAI is simpler but too limited.
>
> But you don't want to say, library data has only poor content?
Poor in terms of the amount of text they carry. A mere title just
doesn't contain many words, and each word occurs in just one form,
much different from full text.
GB has, BTW, the search options inauthor:"name", intitle:"phrase" and
subject:"lc term" .
These are made use of by the LCSH browser which links over into
GB, Worldcat, Librarything and others, but not, ironically, LC:
http://www.biblio.tu-bs.de/db/lcsh/
B.Eversberg
Received on Wed Mar 04 2009 - 05:03:40 EST