Implementasi Metode K-Means dalam Analisis Pengelompokkan Penerimaan Program Indonesia Pintar Berdasarkan Profil Sosial Ekonomi Siswa

Abstract

Dapodik data contains various socioeconomic details about students that can be used to understand their circumstances, such as parental/guardian income, number of siblings, KIP (Student Assistance Card) ownership, KPS (Social Protection Program) enrollment, PIP (Student Assistance Program) eligibility status, and the reasons for PIP eligibility. This data is generally still used for administrative purposes and has not yet been widely utilized to analyze and identify student groups based on socioeconomic characteristics. This study aims to identify student groups and analyze the distribution of PIP proposals at SDN Mangunharjo 7 Probolinggo using the K-Means algorithm. The data used consists of Dapodik data for the 2025/2026 academic year, covering 313 students. The attributes used in the clustering process include family income, the number of siblings receiving KPS, and KIP recipients. This study involved several stages, namely data pre processing, attribute transformation, data normalization, determining the number of clusters using the Silhouette coefficient, implementing the K-Means algorithm, and testing the association and correlation between attributes and PIP proposal status. The results yielded three clusters: Cluster 0 with 252 students, Cluster 1 with 17 students, and Cluster 2 with 42 students. The results of the correlation and association tests showed that all clustering attributes were significant in relation to PIP proposal status. This study demonstrates that K-Means can be used to describe patterns in students’ socioeconomic profiles, but not to determine eligibility for PIP.