Mining Telecommunications Customer Contacts Ring

In recent years, the telecommunication operators have accumulated millions of data on their customers and services. Competition is so fierce that all the companies put more and more money to the CRM system. With the help of CRM, the companies can get the profile of their customers, find new market, and design new services.Data mining is a powerful tool to know our customers and lift their values. At present, the data mainly focus on the characters of our customers themselves, such as their income, job, and age. As we look deep into the data, we can find that the CDR (call data record) also shows us the relationship between customers. If we want to know more about our target customers, we can turn to the customers around them. So, the question is how to get and evaluate the relationship between customers.In this paper, we introduce the basic concepts and techniques on data mining, and the ideas and algorithm of clustering. We also describe the characters about telephone call between customers, and evaluate the relation tightness between them. At the end of this paper, we design an algorithm to mine the relationship cycle of a customer. The algorithm borrows the ideas of Google Page Rank, the ideas of DBSCAN, OPTICS CHAMELEON, and CURE. When the customers are clustered to hundreds of relationship cycles, we try to find the core members and the child ones. Now, we can put our resources to the core customers, and with the help of their relationship cycles, our products and services can be advertised to the customers around the core ones.