Implementasi Sistem Penunjang Keputusan Berbasis Web untuk Menentukan Calon Penerima Bantuan Sosial Menggunakan Algoritma K-Means

Studi Kasus di RW 007 Desa Mekarsari

Abstract

Poverty remains a major social issue, while the determination of social assistance beneficiaries at the neighborhood level may still rely on manual assessment and incomplete information. This study aims to implement a web-based Decision Support System (DSS) using K-Means Clustering to assist administrators of RW 007, Mekarsari Village, in grouping residents according to socioeconomic characteristics and identifying prospective beneficiaries. The study employed literature review, observation, interviews, and questionnaires. The system was developed using the Waterfall model. Resident data were represented by eight criteria: income, occupation, number of dependents, vehicle ownership, marital status, housing status, health condition, and assets. The criteria were converted into numerical values, normalized using Min-Max normalization, and clustered with K-Means. The Elbow Method was used to determine four clusters. Processing 81 resident records produced four clusters, with Cluster 2 containing 15 residents (18.52%) and being interpreted as the eligible group based on centroid characteristics. The clustering process converged after eight iterations, with total SSE of 3.9871 and an average Silhouette Score of 0.2342. Black Box testing of 18 functions showed that all tested functions operated as expected. The system therefore provides a structured data-based aid for RW-level decision making.