<![CDATA[This study examines the effectiveness of the Naïve Bayes and C4.5 algorithms in analyzing book borrowing patterns at the Pringsewu Muhammadiyah University Library. As libraries increasingly serve as vital educational hubs, understanding user borrowing behavior is essential for effective collection management and service enhancement. The research follows the Cross-Industry Standard Process for Data Mining (CRISP-DM), which includes stages of business understanding, data understanding, preparation, modeling, evaluation, and implementation. A dataset consisting of 5,586 records and ten attributes related to book lending was utilized, with comprehensive data cleaning and preprocessing conducted. The performance of both algorithms was assessed using K-fold cross-validation, yielding an accuracy of 96.26% for C4.5, compared to 91.44% for Naïve Bayes. These results demonstrate that C4.5 is more adept at capturing complex relationships within the data, providing deeper insights into user preferences and enhancing library services. This research underscores the potential of data mining techniques to optimize library management and proposes avenues for future investigation, such as exploring advanced machine learning algorithms and expanding datasets for use in broader library contexts.]]>
Copyrights © 2025
