Perbandingan Metode Klasifikasi C4.5 dan Naïve Bayes untuk Mengukur Kepuasan Pelanggan

Data mining is a process of collecting meaningful data from some large information contained in information bases, information warehouses or other documentation places. Classification is an educated educational process (supervised learning). The universal classification procedures used include: Decision tree, K-Nearest Neighbor, Naïve Bayes, Neural Network, C4. 5 as well as SVM. Consideration of the classification method is tried to ensure the type of classification that shares the highest accuracy value of an object. The information used in this research is information from a questionnaire about customer satisfaction with the services of PT. Media Semesta Solutions, the factors used in this research are Reliability, Responsiveness, Assurance, Empathy, and Tangibility. Based on the results of the comparative analysis, the information mining classification procedure is C4.5 and Naïve Bayes proves that the C4 method. 5 is more accurate than the Naïve Bayes method, this result is seen from the accuracy value where the C4 procedure. 5 has an accuracy value of 94, 17%, greater than Naïve Bayes with an accuracy value of 85.83%