Implementation of Data Mining for Classifying Electricity Subsidy Recipients in Stabat District Using the Naive Bayes Algorithm

Abstract

The electricity subsidy assistance program is a government initiative designed to enable low-income households to access electricity at reduced rates. Although the national data system is well-integrated, the economic circumstances of the population are dynamic. Such changes often lead to discrepancies between centrally recorded data and the actual situation on the ground, creating a risk that subsidies may not reach the intended recipients. Without an objective preliminary screening system, data submissions risk rejection and the accumulation of backlogged paperwork. This study aims to implement a web-based application serving as a local early screening tool to classify eligibility by automatically, rapidly, and measurably ranking the priority of prospective subsidy recipients; it employs a quantitative approach utilizing data mining techniques and the Naive Bayes algorithm. The Waterfall method was applied for software development. The results demonstrate that the developed system is capable of classifying the eligibility status of electricity subsidy recipients, specifically for 900 VA household customers. This research is expected to assist relevant agencies in identifying the household electricity customers entitled to receive the assistance.