Analysis of Book Borrowing Patterns Using the Apriori Algorithm (Case Study: East Java Provincial Library and Archives Service)

Abstract

This study analyzes book borrowing patterns at the East Java Provincial Library and Archives Service using the Apriori algorithm. Data from borrowing transactions between January 16 and March 10, 2023, were analyzed to identify relationships between book titles. This data mining study involves data preprocessing, checking for missing values and duplicates, as well as transforming data into a binary matrix. The analysis results show that item combinations with a minimum support value of 1.2% can be identified. Association rules such as AC, AE → AG have a support value of 0.012 and a confidence level of 40.9%. Further analysis reveals more complex patterns, such as AE, AJ → AC, with a confidence level of 57.1%. This research contributes to understanding book borrowing behavior and offers solutions to improve library management efficiency, such as book recommendation systems and more appropriate collection arrangement strategies. Thus, libraries can develop services that are more responsive to community needs.