Karakterisasi Pelanggan PLN Menggunakan Algoritma Fuzzy C Mean

Abstract

Based on the technology point of view, CRM is a system tool in analyzing customers’ behavior based on customer histories data. One of the important tasks of CRM is identifying PLN customers’ behavior based on their behavior in using electricity. Customer identification is done by segmenting customers. In this research, the customer that will be segmented are home, business and industrial customer. The variable that used as a attribute in segmentation process are the duration of subscribing, total payment, discipline in paying and total electric usage. The system that developed in this research provided many choice of variable combination to be segmented. The goal is to provide many choice for PLN in segmenting customer. The method used for segmenting customer is fuzzy c-mean algorithm. To measure the segmentation accuration, the result of the clustering is validated by the Xie and Beni index method. The result of the research showed that business and home customer are more accurate if they are clustered into 3 clusters, while the industrial customer in some variable combination are more accurate if they are clustered into 3 clusters and the some variable combination are more accurate if they are clustered into 4 clusters. The result of customers’ character segmentation can be used as supporting data in making PLN’s business decision which is related to CRM.