This study examined the use of decision tree algorithms, specifically C4.5, in classifying the eligibility for social assistance fund distribution in Kelurahan Bangka Belitung Laut, Kecamatan Pontianak Tenggara. Inaccuracies in distribution were caused by recipient selection based on village officials' recommendations, which only considered the type of occupation. To address this issue, this study developed a system based on the C4.5 algorithm to process economic census data. This research used a quantitative method with a descriptive analytical approach, collecting data through literature studies and economic census, and processing data using the Rapid Miner application. The classification model generated was evaluated using 10-fold cross-validation to ensure high accuracy. The results showed that the C4.5 algorithm achieved 100% accuracy, precision, and recall. The decision tree model indicated that the main attributes determining eligibility were occupation and income. Some rules derived from this model, such as those who are unemployed with an income below Rp29,500 being eligible for assistance, provided clear guidelines for policymakers. The implementation of this algorithm is expected to improve fairness and effectiveness in the distribution of social assistance funds in Kelurahan Bangka Belitung Laut, reduce public dissatisfaction, and prevent potential social conflicts. This study recommended adopting the model in other areas with adjustments to local data to enhance broader fairness in aid distribution.
Copyrights © 2024