Metadata Moves a Library

Library metadata does more than connect researchers to library resources.  It directs users to a specific building, to a specific floor, and leads them to a specific book.  With the renovation of Lilly Library fast approaching, Duke University Libraries Technical Services began exploring ways to leverage library metadata to facilitate the move of the collection from one location to multiple temporary locations.

For the Technical Services Team, preparing the Lilly Library collection for relocation translated into assessing its level of readiness, identifying and addressing gaps, remapping the collection to temporary quarters, and facilitating the return of the collection to its newly renovated spaces.

Addressing the Gaps

The good news is that the metadata, linking patrons to the Lilly Library collection, was in remarkably better shape than that of previous renovation projects.  Much of the credit for this state of readiness goes to the Lilly Team.  While there were a few issues that required our attention, the even better news is that much of this work could be automated.

Smart Barcoding

One of the questions to be answered was, “how many items in the collection lack a barcode?”  A review of an Aleph report revealed that only 6,000 volumes were unbarcoded.  Employing the use of smart barcodes would provide the most efficient means of addressing this task.  Unlike the generic barcode used at the point of cataloging, a smart barcode label contains a call number and a title. The application of the barcode is done by hand and the records in our ILS (Aleph) are updated in batch using an automated process.

With an Aleph report of unbarcoded volumes as our starting point, we contracted the services of Watson Label Products to produce the labels and supply a file with smart barcode numbers matched to our system-generated (“Dummy”) barcodes.  Once the file was received, we worked with Library Systems and Integration Support (LSIS) to develop a script to replace the system-generated barcodes with the smart barcodes.  The barcode labels arrived in sheets of 150 barcodes each, arranged in call number order, with bibliographic information to facilitate matching by the Lilly Team.

Smart Barcode Data File

smart barcode data file

 Spine Label Conflicts

In the process of organizing the data file for smart barcodes, another issue emerged.  A small group of Dewey call numbers appeared in the dataset for a collection that has no books labeled in Dewey.  The “In-Process LC” status in these records pointed to the LC Reclassification Project as the culprit.  From 2004-2008, DUL embarked on a project to convert 2.3 million volumes from Dewey to LC classification.  An automated process designed to copy the vendor-supplied LC call number from the bibliographic record to the holdings record had failed.

Typically, these problems are sent to Technical Services for resolution one-by-one.  At this juncture, was an automated approach even possible?  Some bibliographic records can contain more than one LC call number.  How could we proceed with confidence that the one selected in an automated process is an exact match to the spine label already on the book?  After confirming that every record with an “In-Process LC” status also contained a vendor-supplied LC call number, we turned to the Lilly Team to check the spine labels.  With their verification in hand, the next step was easy.  We used an existing Aleph Service to fix these spine label conflicts.

The method employed to solve this problem serves as a proof-of-concept that can be applied to all DUL collections that may have legacy Dewey call numbers in their holding records.

Weeding the Collection

The Lilly Team identified approximately 2,000 duplicates to remove from its collection.  Out of a desire to retain the copy in the best physical condition, they began the weeding process by retrieving and visually inspecting each copy.  The initial review of 160 books resulted in 80 to be withdrawn.  A file of the 80 titles was sent to Technical Services for records maintenance and the books were sent to Collection Strategy and Development for disposition.

Our Metadata & Discovery Strategy (MADS)  team suspected that library metadata could inform decisions about which copy to withdraw.  Would the number of times a book circulated correspond to its physical condition?   Could the manual selections made by the Lilly Team be replicated in the data?  To answer these questions, we produced a report of circulation history for the 80 pairs of duplicates reviewed by the Lilly Team.  MADS found that in 90% of the cases the items with the most circulations showed the greatest level of wear.

Remapping the Collection

From our perspective, we are not only planning on how to assist in moving the collections out of Lilly; we also have to focus on how to get it back post-renovation.  When we talk about the “Lilly move” we are really referring to the “Lilly moves” because not all materials are being temporarily located in the same space and they are all coming back.  Some materials will be temporarily at the Library Service Center (LSC) off-site repository, some in the Perkins stacks, current periodicals, or the East Asian Collection reference shelves; while others will be in a temporary space on East Campus that we’re calling the Satellite location for now. Immediately it became apparent that temporary location codes were needed for both moving the collections and controlling the display for patrons.

Cleaning-Up Collection Codes

In the process of setting up temporary codes for the Lilly collections, one of the earliest activities was defining what the Lilly Collections were, from a metadata perspective. Based on the collection codes in Aleph, we identified the number of materials in each code.  We discovered some codes were unused and had no items.  We also found a need to collapse multiple codes that served the same purpose, virtually transferring materials from one collection to more appropriate ones.

Part of that work was collaborating with members of the Lilly Team to collapse un-needed collection codes to the more correct codes.  In our clean-up we found out that there were codes that didn’t serve their intended purpose, so we removed them after updating item metadata.  After that, Metadata & Discovery Strategy Department (MADS) staff created temporary codes, determined how they should appear to patrons, and began changing metadata to the temporary codes so that the materials can be shipped. We worked with the Assessment & User Experience Department to adjust some of the public displays after the Lilly Team determined what those should be.  Once the building is empty, MADS staff will have updated all of the collection codes to temporary collection codes with batch processes.  This initial clean-up sets the stage for a more complete inventory.

There was some dissonance between intellectual collections, like the graphic novels, and the actual collection codes in use.  For example, we think of the graphic novel collection as a separate entity from the rest of the Lilly stacks, but there is no separate collection code (or shelving) for the graphic novels.  We just know they are there.  So, to get the metadata into shape so that patrons can see where things are during the renovation, and to be able to identify each collection later, temporary location codes were created for each sub-part of what we know as “the Lilly Collection.”

Throughout Technical Services, there have been a myriad of other changes to workflows in order to support the operations of the Lilly Library as it undergoes renovation.  Some of these changes are temporary, but others are a permanent reflection of our adaptability and flexibility.

Remember the first sentence in this post: “Library metadata does more than connect researchers to library resources?”  The Lilly Library renovation can now proceed with confidence that patrons can still explore its rich resources with ease.