Automated Metadata: Transforming ETD Cataloging at Oklahoma State University Library
In the realm of cataloging Electronic Theses and Dissertations (ETD) whether traditionally one by one or in batches, the challenge of finding a balance between the quality and the quantity of cataloging is still a universal challenge among academic libraries, regardless of their size. The emphasis on the quality of the cataloging has never been more significant as ETDs, once given minimal level description in library catalog records, now are rich with descriptive metadata as they come to represent an important role of scholarly communication for universities in both their catalogs and institutional repositories. In 2014, the Oklahoma State University Library addressed this challenge with sweeping changes to the cataloging of its ETDs. The transformation began with reducing the amount of time on cataloging the same information for MARC records entering the library’s catalog and for DSpace records of its’ newly launched joint institutional repository, ShareOK. At the core was a batch process to expedite ETD cataloging for both processes while ensuring that the cataloging was meeting national standards. Utilizing Excel to extract data from the ProQuest XML files allowed the use of built-in Excel formulas to transform the data into a standardized form. MARC fields were added and edited to meet RDA guidelines and OCLC record requirements. In addition, Dublin Core elements were standardized in the DSpace records so that both records conformed to national standards for quality and consistency in data. With basic editing and reviewing for quality control, the transformed data is exported as Dublin Core elements for DSpace and also mapped and imported into MarcEdit to create a MARC record. This new process significantly reduces the time spent in creating, editing, and reviewing while improving accuracy.