<![CDATA[This study aims to design and develop a web-based Customer Intelligence system that integrates customer segmentation, customer value prediction, and location mapping into a unified platform. The Fuzzy C-Means algorithm is used for customer segmentation based on behavioral characteristics, while the Decision Tree Regressor is applied to predict Average Revenue per User (ARPU). In addition, the system incorporates Point of Interest (POI) visualization to support location-based analysis. The results show that the clustering model successfully identifies three to four customer segments with distinct characteristics based on ARPU, service usage, and customer loyalty. The prediction model achieves a good performance with an R² value of 0.8905 on the testing data. The system is also able to automatically generate service recommendations based on customer segmentation results. Therefore, the proposed system can assist telecommunication companies in formulating more effective and data-driven service strategies.]]>
Copyrights © 2026
