From Citations to Decisions: AI-driven Citation Analysis for Transforming Library Collection Strategies

Abstract

Libraries face growing challenges in meeting evolving user needs due to constrained collection budgets and rapidly changing technologies. As a result, librarians must be more strategic in their collection development and assessment practices to determine which materials best meet user needs. Standardized metrics such as usage, search, and turnaways counters are used to determine critical decisions, but often provide limited insights. More robust approaches, such as citation analysis, can provide a deeper understanding but are often labor-intensive and difficult to scale. Using ten years of citation data, this paper presents an AI-assisted methodology for collection management. The approach identifies not only which resources are used, but also trends in how and what types of materials are cited in a discipline. By integrating AI into the analytical workflow, the method significantly accelerates processing while maintaining high accuracy (97.16%). The project demonstrates how libraries can strategically adopt AI to advance practical, scalable, evidence-based decision-making. This study also highlights a way to transform budgetary and technological challenges into opportunities for innovation, enabling libraries to manage collections more effectively to meet user needs and to advance their evolving role in research support.