Implementation of Data Mining Using K-medoids Clustering Method for Determining Social Assistance Recipients

Nurzainun, Ratnasari
Sinawati, S
Prayogi, Denis

IJISTECH
Vol 8, No 5 (2025): The February edition

Regarding social assistance, the village head, “Lurah” of Selumit, is currently reviewing data on residents based on government-provided statistics on the capabilities of low-income families and the need for social assistance.  Therefore, this study proposes K-medoid clustering to ensure that the assistance provided is appropriate. The study collected 62 data on social assistance recipients consisting of 6 criteria, namely employment, assets, income, jak (who are still dependents), home status, and home conditions. K-Medoids analysis using the Euclidean distance function with K=3 produces cluster 1 with 12 data, cluster 2 with 31 data, and cluster 3 with 19 data. The recipients prioritized to receive social assistance are the data in cluster 2 by calculating the average of the most considerable maximum weight value.