One of the items that's been on the back burner for a while is working with OCLC's Bibliographic Record Notification Service. BRNS, which is included in OCLC's subscription pricing, supplies a daily file of MARC records with the Library's holding symbol attached that were updated during the previous day. OCLC provides a variety of options for how inclusive the files should be. Ours at McMaster are limited to monographic items (serials and other continuing resources just get way too complicated, and our e-resources Librarian is also working on the MARCit service, so we don't want competing record loads...), but is otherwise pretty open. We could limit it to certain kinds of updates, but before doing that I wanted to see what we would get if they just sent everything.

Having now worked with several hundred of these records over the past few weeks, some patterns are emerging:

Changes to basic descriptive elements are fairly rare--Perhaps not surprising. The title is the title, after all, unless there was a typo or other error in the original record (which hopefully someone caught when it was loaded). I am occasionally seeing more complete pagination information in the new record, sometimes the publisher's name is slightly fuller, or maybe they've added another publisher location but, by-and-large, nothing major.

Many of our older records are missing the item dimensions, and this addition has been more frequent. Omitting the dimensions wasn't really considered significant until we recently tried using the data to support our mass digitization program. Oops. Just one more reminder that what seems trivial today may turn out to be important later.

More extensive subject headings and notes, and the addition of ISBNs and classification numbers are among the other enhancements to our older records.

Indicators, subfields, and 00x coding--Changes here have been more frequent simply because we have so many records with now-obsolete coding in the catalogue. Again, this is one that's easy to initially dismiss as trivial. For years our OPACs have largely ignored subfields, indicators, and such, passively displaying the contents of catalogue records with little regard for how the information is coded. With Endeca powering our front-facing catalogue at McMaster, we're now using subfield codes, as well as codes in the 00x fields and a variety of others, to drive faceted browsing. Getting updated records into the system therefore has a direct impact on the user experience by improving the consistency and completeness of the facets.

On the other hand--Sometimes the change IS dramatic. I've seen a number of records like this,

LDR 00777Cam 22001817? 45
003 BIBL
005 19940517000000.0
008 940405s ||||onc|||||b|||||||||||eng|d
050 14 $a LB
110 10 $a ONTARIO. $b MINISTRY OF EDUCATION. $b RESEARCH AND EVALUATION BRANCH
245 10 $a SCHOOL ATTENDANCE AND NON-ATTENDANCE IN CANADA AND THE UNITED STATES:
          CL ASSIFICATION OF MALE YOUNG OFFENDERS WITH SCHOOL ATTENDANCE PROBLEMS.
265 $a (7)
300 $a Microfiche
440  0 $a ONTARIO MINISTRY OF EDUCATION RESEARCH AND EVALUATION BRANCH RESEARCH REPORTS. $v 4702
700 10 $a PAUKER, J.D.

which have also been sitting in our catalogue for years, transition instantly to full cataloguing, complete with subject headings and an LC class number in the 050 (again, used for browsing in Endeca).

Content is king--By far the most common update is the addition of contents notes. Even records for items published in the early 20th century are getting this treatment. Certainly not information we would get but for someone at another library deciding it was important for their users and being willing to share with the rest of us. Having the contents listed in the MARC record itself also has the benefit of making them searchable.

It's relatively painless--Records with OCLC numbers overlay seamlessly with no intervention. Our standard record loader protects certain fields (local notes and such), maintaining them in the updated record. Older records added to the catalogue before we joined OCLC frequently have to be manually merged with the incoming record. But, as these are generally the records most in need of updating, it's a worthwhile exercise. It's also not as time-consuming as it might at first appear. I've found that I can process a week's worth of files in a couple of hours, frequently less.

Overall, I've found BRNS to be a useful tool for improving the catalogue. We're not touching every record for every item, just those where an improvement is available. The changes being made to existing records are not undoing or redoing the work of our cataloguers, but adding to it. And finally, most of the updates are not "things only a cataloguer could love" (and I say that as someone who appreciates the nuances of cataloguing), but have direct benefits to the user's experience.

OR IMMEDIATE RELEASE

Contact: Tommy Stringer

864-918-4859

Tommy@ElectStringer.com