Projects

Contents

Merging LP records formerly under AACR1 cataloging practices

Merging LP records formerly under AACR1 cataloging practices

Under AACR1, each musical work on an LP sound recording was often cataloged on a separate bibliographic record, leading to multiple bibliographic records describing one item. The linkage between the records for a single item was in the descriptive metadata (usually in a 501 MARC field). This was a similar practice to “bound-withs” in print media. When relational databases came into being, though print “bound-withs” were often linked by an item record associated with multiple holdings records, the same was not true for sound. This already makes it difficult to pull together all of the descriptive metadata for a single item, as library catalog users must begin an entirely new search to find the rest of the content description.

During my time at SMU I embarked on a project to run a complete inventory of the LP collection, part of which included reconciling these multiple records into one. This was a fairly straightforward task, as the collection was relatively small, an accession-based call numbering scheme was used regardless of the source, and there were very few instances of multiple copies of one recording. Automation (e.g. adding barcodes when the online database was created) was completed in its entirety, so each item on the shelf had a barcode. As such, in a spreadsheet that listed the entire collection, it was easy to pinpoint the extraneous records, as they did not have an associated barcode. Though there were some exceptions which inevitably turned out to be errors, the process was easy to develop and implement.

I undertook the same project at IU, which proved to be a much more complex and involved process. First, automation and retrospective conversion were never completed, leaving approximately 30-40% of the collection out of the online database and without barcodes. Second, a locally-developed, topical-based call number scheme was used for most collections, and an accession-based call number scheme was used for others. Third, major collections given by a donor added a prefix to the associated call numbers, leading to multiple unique call numbers under a given title when a copy was held in multiple collections. Fourth, IU Libraries system chose not to utilize MARC Holdings records. All of these factors made it impossible to easily pinpoint and collocate the relevant records for each item. On top of that, local practices of adding additional copies were inconsistent at best, and therefore a majority of cases required unique and specialized treatment.

TIMELINE: 7/2013-2/2014 (SMU); 2/18/2016-present (IU)

STATISTICS: Approx. 450 items processed (SMU); (IU)

 

COMING SOON:

John Mack Collection Push

IUIE Reports, then Library IT report

Cleaning data and fitting it into the needs of the MDPI workflow manager

Recleaning data

finding issues with separately-cataloged sides, internal dups as well as dups with other collections already going to MDPI

Led to list of several potential projects and catalog fixes/decisions that are desperately needed

COMING SOON:

IU Performances

Started with look at how the MARC data is crosswalked to Avalon

Decided it would be beneficial to MCO users if the titles of all of the performances weren’t identical, and would also be beneficial to all catalog users

Decided to start project to recon every program with a potentially abstract title and make them into more meaningful titles

Pulled IUIE reports of all IUPERFs

Tested several categories for the kinds of changes to make

Ironed out the process we would take, divided between me, Janet, and Emma

Discussed every possible permutation of the changes we could find, and came to a consensus on the format that was the most clear and meaningful