Klasifikasi Bidang Masalah Pengaduan Masyarakat Pada Dashboard “SIGAP” Menggunakan Metode Support Vector Machine

Abstract

The People's Representative Council of the Republic of Indonesia (DPR RI) is an institution that absorbs, collects, accommodates and discusses the aspirations of the people. Therefore, the Secretariat General of the DPR RI requires a dashboard visualization data that can be used to support this goal, namely by making a dashboard. An intelligent system is needed that is able to automatically classify according to the problem in the complaint on the dashboard. Based on these problems, this research was conducted using the Support Vector Machine (SVM) for the classification of the date of care in accordance with the existing problem areas. The SVM method consists of a training process as system learning and testing to obtain classification results. The parameters of the tests carried out are testing lambda, complexity, and maximum iteration. This research uses the Support Vector Machine (SVM) method, which is a learning system. The data used in the study were 1,299 data which were divided into 6 classes. This research class represents 6 classes, namely Law, Defense and Agrarian Reform, Manpower, Education, Energy Resources and Minerals, Health. The SVM algorithm is a linear classification method, so it uses a kernel to deal with nonlinear data. The final result of this research produces the highest average accuracy of 70% C = 10. These results indicate that the model can predict new public complaints with fairly good accuracy.

Keywords: DPR RI, support vector machine, classification