[Koha] Advice on MARC21 fields

[Steven F.Baljkas]

Thursday, February 16, 2006     13:16 CST

Greetings, Rodica,

Sorry for the delay in responding to your message. Below you will find response to your original question/s as well as some advice on cataloguing matters I felt were related and worthy of your attention at this phase of your Koha automation.

Since you are using MARC21 records, your most likely sources will be North American libraries, particularly the Library of Congress. MIT would also probably be a very good source.

Given that many books published in the U.S.A. are simultaneously published in Canada, Canada's national library also has them in our national union catalogue, which may facilitate cataloguing for you at your institute.

The reason for that is that, unlike LC, Canada's national library allows for batch searching for individual records via its free (registered) AMICUS service. That is, you can search for each record individually, but keep the results for download and further editing/adjustment in batches.

Although some Koha-ites would recommend using computer scripts to locate books by ISBNs, as a cataloguing purist I would continue to hold that you are best served by ensuring that you have found a correct and quality record for each title in your collection.

AMICUS helps with that by allowing you to see the actual MARC cataloguing copy before ultimate selection of the record. Using a script, at best, you may have the first though not often the best record. At worst, you may get false negative returns (i.e. if the ISBN has been omitted or miscoded for some reason -- and it does happen more often than people would choose to admit freely -- the search will say the book record does not exist when in fact it does and only requires that field be added) and false positive ones (where an incorrect edition or even incorrect work is retrieved: and yes, it does happen that the supposedly unique ISBNs are accidentally assigned to different works).

In any case, that is my little (partially self-serving but genuinely sincere) promotion for 'traditional' catalogue record searching.

As for AMICUS, if you want to find out more about how it might help you at IMAR, you can surf to 
  URL <http://www.collectionscanada.ca/amicus/006002-122-e.html>
to learn about it and sign up if you so choose. (Regarding the requirement to sign-up, I can tell you honestly that in the several years I have belonged, they have never bothered me with any unwanted mailings; in fact, the only contact has been to advise of service upgrade downtimes and otherwise unscheduled service downtimes, and that only 2 times so far).

In terms of cataloguing issues, I should mention that with mathematical works -- and scientific monographs in general -- ***you need to consider carefully whether you will be using the LC subject headings (LCSH) or a specialised thesaurus.*** It is likely that LC and other libraries (CISTI, MIT, e.g.) have MARC cataloguing your works; it may be much less likely that they have used anything other than LCSH for subject analysis. Although the terminologies may be similar, you may find it better to structure your own thesaurus for local use.

Given Koha's present problems with understanding any subfields beyond the initial 650$a properly (and Paul Poulain's suggested work around is not valid: it violates MARC standards by including perforce two $a's - a Non-Repeating Subfield -- in a given 650), ***you may want to make use of your time to add to your records keyword terms that will be meaningful and useful to your library patrons and have those linked to Koha for searchability***. This is just a suggestion but you may find it helpful in the long run.

Finally, while it is important to configure Koha carefully (MARC on/off, MARC21 or UNIMARC flavour, and so on), I was unsure what you meant about "control fields".

***At present, Koha does not make use of what are commonly called control fields (in cataloguing terminology, Leader to 008 inclusive). That is something still coming in the future. As things stand now, any info in those fields is effectively lost to Koha.***

In terms of things you would want to configure carefully, though, if you are using the visually appealing NPL template of Koha, be aware that you will need to set up your ItemTypes very carefully, decide on/design appropriate icons for the graphic display, and ***code 942 fields in each MARC record as appropriate for the item in hand.*** That will add to your cataloguing work time but you may find it well work the extra effort.

I hope that this information has helped somewhat. If you would like me to run a test of AMICUS to see what kind of success it would have in retrieving suitable records for you, please feel free to contact me off listserv so we can set up a test base.

In the meantime, welcome to the Koha community and best of luck with your efforts in automating your library at IMAR using Koha!

Cheers,
Steven F. Baljkas
library tech at large
Koha neophyte
Winnipeg, MB, Canada

============================================================
From: Balasa Rodica <rodica.balasa at nemesisit.ro>
Date: 2006/02/09 Thu AM 08:00:12 CST
To: koha at lists.katipo.co.nz
Subject: [Koha] Advice on MARC21 fields

Hello everybody,

I write hoping that you could help us with advice about MARC21
configuration. 

My name is Rodica Balasa, Project Manager at Nemesis IT
from Bucharest, Romania. We are helping IMAR (Romanian Institute of
Mathematics) to install Koha for library management. Currently the
library catalog is recorded in writing (no database), so it is important
that the MARC description of most of the books to be downloaded from the
z39.50 public servers.

We arrived at the point of configuring the MARC21 control fields, and as
I understand that the choices we make are very important, I thought of
asking for advice other libraries that have more experience with Koha.

Thank you very much!

Rodica Balasa

______________________________________________
Koha mailing list
Koha at lists.katipo.co.nz
http://lists.katipo.co.nz/mailman/listinfo/koha
============================================================