<![CDATA[The library is an essential source of information that plays a vital role in supporting students’ learning processes. However, the increasing number of book borrowing transactions often makes it difficult for librarians to recognize borrowing patterns and manage book collections and acquisitions effectively. This study aims to discover borrowing patterns in the library of SMP Negeri 2 Denpasar using the Knowledge Discovery in Databases (KDD) method. The research process consists of five stages: Data Selection, Preprocessing/Cleaning, Data Transformation, Data Mining, and Interpretation/Evaluation. In the Data Mining stage, the Apriori algorithm was applied to generate association rules from 3.004 borrowing transactions. The results revealed two dominant patterns: students who borrowed the Pure Science category also tended to borrow the Social Science category with a confidence value of 53%, support of 14%, and lift of 1.87, while those who borrowed the Art category tended to borrow the Literature category with a confidence of 71%, support of 36%, and lift of 1.33. These findings demonstrate that the KDD process combined with the Apriori algorithm can effectively identify borrowing patterns and provide insights for optimizing library collections and management. Based on these patterns, this study recommends arranging the bookshelves according to categories that are frequently borrowed together and increasing the collection in categories that show high levels of student interest in the association results.]]>
