NASIG 2026

Academic libraries steward large and complex collections to support teaching and research, yet determining how well these collections align with what users produce, seek, and use remains a persistent challenge. Traditional assessment practices rely primarily on circulation and e-resource usage metrics—measures that capture only a fragment of user engagement. These approaches overlook two equally important dimensions: faculty research output, which signals institutional scholarly activity, and user search behaviour, which reflects articulated information needs. This project introduces an integrated framework for collection intelligence: an AI-enabled model that brings together three key data sources—faculty research outputs, user discovery behaviour, and item use—within a unified analytical structure. Using Google Gemini Pro, each dataset is classified with the Library of Congress Classification (LCC) scheme and then analyzed to identify patterns of subject alignment, unmet demand, and emerging areas of scholarly interest. Viewed collectively, these three lenses offer a richer understanding of how library collections correspond to institutional knowledge production and user behaviour. The project moves libraries beyond siloed metrics toward integrated, evidence-based insight. It delivers a scalable, replicable model in which AI serves not only as a classifier but as a connector, linking research activity, discovery practices, and actual use to surface the library’s evolving role in the research ecosystem.