Re: OCLC annoucement

From: Charley Pennell <cpennell_at_nyob>
Date: Thu, 23 Apr 2009 18:38:30 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
Hi Marc-

  A lot of libraries, including my own, have been queuing for OCLC's 
"reclamation service" by which we atone for all those years of using 
usurper services such as LC's Z39.50 port, or DRA's, or book vendor 
services, and send an extract of the entirety of our MARC data to the 
utility for matching against WorldCat.  In return, OCLC sends you back 
lists of your local control number and the corresponding OCLC#.  We just 
used API to add the latter to 035 (OCoLC), which retrospectively gave us 
missing OCLC numbers along with updated numbers to cover superseded 
ones.  At the end of the process, we had only about 1% of titles w/o the 
ubiquitous numbers, a high proportion of which were reserve materials 
that should never have been sent but had the wrong item type/location 
attached.  So, getting the OCLC numbers is not so problematic, and if 
your ILS enables APIs, adding them to the catalog is not problematic 
either.  I know that you are also Unicorn people, so you should be good 
to go.

  We used the reclamation service to enable us to roll up display based 
on a common unique ID with our TRLN partners in TRLN Endeca 
(http://search.trln.org) and soon for our UNC-wide union catalog.  Of 
course there is no guarantee that we are all using the same OCLC record, 
as WorldCat has many hundreds of thousands of cases where multiple 
records describe the same manifestation.  Still, the OCLC number is 
about as close as we're going to get to a unique URI for the objects 
described in most of our catalogs.  He who has the best available 
metadata (not the best possible, mind you) wins, whether we like it or 
not.  I might have hoped that LC could have become a true national 
library to fulfill the "best metadata" role as they did in the print 
union catalog era, but alas the "not for profit", but essentially 
private, sector has won out.  Now we will see what ransom must be paid.  
I hope that this isn't the death knell for the "blooming of a thousand 
flowers" we've seen on the search side over the last few years.

   Charley Pennell
   Principal Cataloger for Metadata
   NCSU Libraries

Truitt, Marc wrote:
> FWIW, we've been considering WCL membership / subscription /
> whatever-you-call-it with OCLC and we have a consortial database of
> several million records, very few of which include OCLC numbers.  OCLC
> is offering what it terms a "remediation" service, wherein it takes a
> dump of our bib records and runs a matching algorithm against it,
> returning to us the OCLC record numbers matched.  We've been given to
> understand that we should expect the overwhelming majority of our
> records to be matched correctly.- mt
Received on Thu Apr 23 2009 - 18:40:10 EDT