Analyzing library users' borrowing behaviors using a data mining approach: The case of çankaya university

This study aims to analyze the borrowing data of Çankaya University Library between 2014 and 2024 using data mining techniques, to reveal the information access habits of its users. In the study, the data set was preprocessed. To address the identified research questions, the K-Modes clustering algorithm was used to group categorical data, and significant patterns and relationships were revealed using Apriori association rules. The results show that the number of borrowed books increased steadily between 2014 and 2019, but a considerable decrease occurred during the COVID-19 pandemic (2020–2021). The Faculty of Engineering (11,216), followed by the Institute of Social Sciences (5960) and the Faculty of Law (5439), has the highest loan rate on a unit basis. In contrast, the Preparatory School (44) and the Department of Foreign Languages (58) have the lowest loan rate. At the departmental level, Law (5977), Private Law (3357), and Industrial Engineering (2327) were the most active. In terms of user profiles, undergraduate students were the most intensive group with 23,350 borrowings. Regarding subject categories, Turkish Law (KKX) and English language resources (EASY/PE) were the most frequently borrowed. Analysis using the Apriori algorithm reveals the most frequent associations between KKX, Master's degree (support: 0.125) and KKX, Law (support: 0.120). Cluster analysis using K-Modes created five different clusters, and a partial separation was observed between these clusters. For engineering data, concentration was observed in clusters 1 and 3, while for law data, concentration was observed in clusters 2 and 4. Overall, the study demonstrates that data mining techniques are effective in understanding user behaviors in academic libraries and can provide significant contributions to collection development, user services, and resource planning.