Perbandingan Kinerja Algoritma Naïve Bayes dan C4.5 pada Sistem Web Klasifikasi Kelayakan PKH

Abstract

This study discusses the development of a web-based classification system for determining the eligibility of recipients of the Family Hope Program (PKH), by comparing two data mining algorithms: C4.5 and Naïve Bayes. The dataset used includes various attributes relevant to eligibility assessment for social assistance. The C4.5 algorithm is employed to generate an interpretable decision tree, while the Naïve Bayes algorithm is used for probabilistic classification. The results show that Naïve Bayes achieved the highest accuracy at 98%, excelling in processing large datasets more efficiently. Meanwhile, C4.5 achieved an accuracy of 93.33% and offered better interpretability through its decision tree visualization. Both algorithms proved effective in classifying PKH eligibility and can be implemented in social assistance information systems to improve the accuracy and efficiency of the beneficiary selection process. This research concludes that the choice of algorithm should be based on system priorities—whether the focus is on processing speed or result interpretability.