<![CDATA[University libraries require systems capable of detecting abnormal book borrowing behavior to maintain service effectiveness. Without an anomaly detection mechanism, libraries are prone to various operational issues, such as increased manual verification workload for librarians, delays in identifying problematic borrowings, and discrepancies in book inventory records. Based on preliminary data prior to this study, an average of 8–12% of transactions exhibited irregularities, such as extreme delays, borrowings outside operational hours, or unusually large numbers of books borrowed in a single transaction. This condition highlights the need for an analytical approach that can provide faster and more consistent detection than manual inspection. This study applies the Isolation Forest algorithm to detect anomalies in the book borrowing dataset of the Wiraraja University Library. The borrowing data are processed through data cleaning, feature extraction, and standardization before being used for model training. The results show that the model achieves an accuracy of 90% in identifying normal borrowing patterns and successfully detects transactions with extreme characteristics as anomalies. These findings confirm that a machine learning–based approach can improve library operational efficiency while minimizing the risk of abnormal transactions that were previously difficult to detect through manual processes.]]>
Copyrights © 2026
