<![CDATA[One of the main challenges faced by libraries is determining the procurement of book collections that align with the needs and interests of borrowers. At the Sragen Regency Archives and Library Service, book procurement is often based on intuition or unstructured requests, resulting in many books that are less popular among visitors. This leads to a low number of book borrowings, meaning the library cannot provide optimal services. Based on this issue, the author attempts to cluster the book borrowing data for the year 2023 from the Sragen Regency Archives and Library Service by age group, book categories, number of borrowings, number of titles, and number of copies using data mining techniques with the k-means clustering algorithm. For the initial data processing, the author uses the Min-Max normalization method. After normalization, k-means is calculated with 3 clusters, followed by finding the optimal cluster using the elbow method, silhouette, and gap statistics. The results of the optimal cluster are compared with the results from the Dunn Index method. The research identifies three clusters: Cluster 1 contains book groups with low interest, consisting of 7 categories: General Works, Social Sciences, Language, Pure Sciences, Applied Sciences, Arts and Sports, History and Geography; Cluster 2 contains book groups with moderate interest, consisting of 2 categories: Philosophy and Psychology, Religion; and Cluster 3 contains the book group with the highest interest, consisting of 1 category: Literature]]>
Copyrights © 2025
