Sistem Klasifikasi Pengaduan Masyarakat Pada BPJS Ketenagakerjaan Menggunakan Algoritma Naïve Bayes Berbasis Mobile

Abstract

Public complaints are needed so that the performance of an agency can be known whether it is running well or vice versa. Improving the quality of public services can be done through the quick resolution of complaints from service providers. Naive Bayes Classifier is an approach that refers to Bayes theorem by combining previous knowledge with new knowledge. So it is called a classification algorithm that is simple but has high accuracy. For this reason, this study will prove the ability of Naive Bayes to classify public complaints against BPJS Employment which contain information on services or social conditions in Pekanbaru City. There are 3 types of complaint classification that can be submitted, namely disbursement problems, BPJS Employment service problems and problems with social security administration. The mobile-based system is implemented with the programming language used, namely Java, while PHP is for administrators. Administrators in this system are employees who work in the BPJS Employment office. This system is tested using whitebox testing for unit testing and integration testing, blackbox testing for validation testing and usability testing. The results of this study are the classification system for public complaints at BPJS Employment using the mobile-based Naive Bayes algorithm. The classification accuracy using the Naïve Bayes algorithm is 0.9, the average precision is 0.93, the recall is 0.91, and the F1-score is 0.9.