Creating a microform resource type assisted by ChatGPT [Zoom session]

San Francisco State University Library houses microforms at multiple locations. Despite their unique research value, they are unlikely to be top choices for most students. There is a need to deprioritize them in Primo VE result ranking. Isolating them for ranking configuration is challenging due to the lack of location-based mapping, evolving cataloging practices over time, and the use of a single MARC record for both digital and microform formats.

This presentation will illustrate how leveraging ChatGPT made the process more manageable both in identifying relevant MARC fields and in generating regular expressions. The presenter will also discuss the solutions in creating sets that match true microform records while excluding online resources. Having a microform resource type not only allows for using the boost factor but also enhances searching and discovery. The AI-assisted approach in creating new resource types can be valuable for attendees’ own library’s practices.