Unlocking insights from academic library data using clustering and recommender dashboard analytics for enhanced book collection management

This study, titled "Unlocking Insights from Academic Library Data using Clustering and Recommender Dashboard Analytics for Enhanced Book Collection Management: UST Perspective," explores data-driven strategies to enhance the University of Santo Tomas (UST) library’s collection management. The K-Means clustering algorithm was used to analyze library collection data, identifying patterns based on book titles, publication years, authors, and categories. The clustering results revealed high-demand clusters, including Social Sciences, Health Sciences, Humanities, and Science and Technology, while low-usage clusters highlighted underutilized resources such as Senior High School (SHS), Heritage, Junior High School (JHS), Education High School, and Music collections. Acquisition patterns showed peaks in specific years and emerging categories, particularly in Science and Technology. Data visualization tools like Tableau and JupyterLab were used to present these insights. Despite challenges, such as handling interdisciplinary overlaps and managing data inconsistencies, the K-Means algorithm effectively uncovered meaningful patterns. To enhance user experience, a personalized recommender system using collaborative filtering was developed. This system provides offered book suggestions based on users’ interests, reviews, and ratings by analyzing similar users’ interactions. The recommender system is accessible via www.ustlibrary.online. This study highlights the potential of combining clustering algorithms and recommender systems to support data-informed decision-making, ultimately fostering a responsive and user-centric academic library service.