A Big Data Approach for Customer Behavior Analysis in Telecommunication Industry

The evolution of telecommunications has led to a profound transformation in the realm of communication, revolutionizing how this industry mines customer behavior for their business outcomes. The analysis of user historical activities promoted paramount importance in driving strategic decision-making to enhance customer experiences and recommend ways to attract customers more effectively. While the demand is growing, some telecom data analytics either use small datasets or provide a high abstract level of analysis result. When the number of customers increases significantly, it becomes impractical to customize service for each customer under the same approach. This paper provides a comprehensive examination of the challenges, needs, and solutions associated with the analysis of user data within the telecom domain. We focus on three key user data analysis problems: user clustering, user classification, and revenue prediction derived from user insights. With Florus - our proposed big data framework, we have carried out the telecom customer behavior analysis with a large dataset. The experiment result demonstrates the promising performance and its potential for long-term use.