[DCRM-L] Info from OCLC quality group group re edits to rare book records

[Shiner, Elaine]

Greetings:

I'd like to respond to one point in Nina's letter (without addressing the question of whether OCLC should be making some of these changes on any records)

Nina writes (under #2):

"... it is this institution's policy to omit adding dcrmb in the 040 since the materials are late 19th- century or early 20th-century."

However, DCRMB states (in I.2, p. 11): "Unlike its predecessors, which were intended to apply exclusively to pre-1801 imprints, DCRM(B) may be used for printed monographs of any age or type of production."

Since it seems that protection from both merging and auto-editing depends entirely on the date and the $$e field, perhaps institutions should re-visit their policies about when to use rare book rules and when to code dcrm(x).  If newer materials require rare treatment for any reason, it's entirely permissible to use dcrm, along with the appropriate codes.

Elaine

Elaine Shiner,
Rare Book Cataloger
Houghton Library



From: dcrm-l-bounces at lib.byu.edu [mailto:dcrm-l-bounces at lib.byu.edu] On Behalf Of Dooley,Jackie
Sent: Friday, March 21, 2014 8:15 PM
To: DCRM-L
Subject: [DCRM-L] Info from OCLC quality group group re edits to rare book records

Greetings--

I've gathered some intelligence about editing of WorldCat records to address concerns raised on this list over the past week. Our colleagues in the quality control group are always happy to clarify things in order to keep the community informed and knowledgable about how this work is done. The quality control group has about a dozen MLS-qualified catalogers who know their stuff.

Macro for auto-editing of various fields: The macro is a zillion lines of code, so not possible to articulate everything it does, but it's a combination of RDAifying and correcting routine common errors. Also stuff like adding the FF country code based on 260$b if it's missing and adding 043 based on subject headings. The script bypasses rb records based on 040$e and/or date (pre-1800). Many elements of OCLC's script can be found in the public documentation about Connexion macros that OCLC writes for catalogers to use: http://www.oclc.org/content/dam/support/connexion/documentation/client/basics/macros/macros.pdf

Expanding abbreviations: This is never done in rb records. For others, they're totally cognizant of the need to have item in hand before making a change. Some catalogers send a PDF to show the original data.

Adding RDA 3xx fields: They do automatically add these to records when there's another reason to "touch" them, whether in response to a member error report or elsewise. The values are taken straight from coded data (leader, fixed fields, 008) and translated into the valid RDA terms that match. The 3xx's have been added to OCLC indexes and so enhance retrieval. If rb catalogers see any terms missing, let OCLC know so they can pursue additions. If RDA terms were changed, OCLC would change records in accordance. In the meantime, is there harm in their presence? For the details, see the link to which Kate Moriarty pointed us the other day: http://www.oclc.org/en-US/rda/new-policy.html

Edits made in response to member change requests: The volume of requests they receive isn't huge (at least by OCLC standards! About 3,000/month)--but some identify a pattern that affects lots of records. For example, the 6.2k error requests received in Jan/Feb 2014 resulted in more than a million edits.

Correcting access points: a really high percentage of the edits made by OCLC are done by machine to bring headings of all kinds into accord with authority records. I'm guessing this wouldn't be controversial to this audience-eh? In addition to ongoing edits to match headings that come in on new inputs, lots are being made to conform to RDA (e.g. spelling out "dept"). It's all spelled out here: http://www.oclc.org/en-US/news/announcements/2012/announcement1.html<https://urldefense.proofpoint.com/v1/url?u=http://www.oclc.org/en-US/news/announcements/2012/announcement1.html&k=AjZjj3dyY74kKL92lieHqQ%3D%3D%0A&r=Epy0%2BbwECk089dbPLkByUrrIgIl9%2BPzyUI6irqGT6UE%3D%0A&m=v4O0Amxnay5ROsXHmhKQmyV5ONFO%2F%2BF766mZ0OmHwzA%3D%0A&s=063b9575d513fc80d8c7ae1b3435d78aaa01bd367c77a7b9e1785676aa0b46a2>

Adding 040$e for gihc and amrem: They'll make this happen. It's not a simple matter of plugging the values into one bit of code; they have to determine which other systems may be affected by the addition, etc etc. So it takes some time.

Happy cataloging! Best wishes, Jackie

--
Jackie Dooley
Program Officer, OCLC Research
Past President (2012-2013)
Society of American Archivists



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserver.lib.byu.edu/pipermail/dcrm-l/attachments/20140325/7a320c68/attachment-0001.html>