Joseph Donnelly
Associate Director of Discovery Services
Colgate University
As concepts and terminology evolve, so must library records to remain discoverable and effective. Colgate University’s Discovery Services team developed a workflow to evaluate the Library’s Primo search analytics against a curated list of preferred terms supplied by Cataloglab.org and their corresponding Library of Congress Subject Headings (LCSH). Lists were checked for personally identifiable information using Gemini, then compared using ChatGPT. The analysis revealed stronger matches in preferred terms, particularly when compared against the Library’s zero-hits list. This led to updating a specific set of records by adding either 650 or 690 fields. Beyond record enhancements, it provided insights into the overall usage model of subjects and how they are reflected in user search methods.
