Artificial Intelligence (AI) has recently significantly transformed metadata workflows in libraries. While chatbots and related services have made AI more accessible, fully integrating AI into workflows remains challenging. This presentation highlights the use of AI at the University of Central Florida (UCF) Libraries to assist in assigning subject headings, including Faceted Application of Subject Terminology (FAST) terms and keywords, to both digital and traditional collections.
UCF Libraries adopted the OpenAI API to generate subject headings, conducted evaluations and comparative studies of AI-generated terms and contributor-supplied terms, revealing the potential benefits of AI in metadata creation. This approach is particularly valuable for libraries with limited staff resources, enabling efficient metadata creation for diverse collections, including unique local materials. Furthermore, AI-generated subjects provide meaningful references for cataloging specialized collections.
While the project’s outcomes are promising, the extent to which AI can be integrated into workflows varies depending on local circumstances. Incorporating AI into workflows introduces new challenges and opportunities for librarians, and collaborating with external