Application of Data Mining in Predicting Cell Phones Subscribers Behavior Employing the Contact Pattern

Following telecommunication services becoming competitive, client contract management in this sector has again much importance. Regarding the fact that a huge volume of telecommunication data especially details of the cell phone conversations exist, and they are practically not used, employment of data mining techniques on such data lead to exploring the hidden knowledge in them on the subscriber´s behavior and lead s to predicting their behavior. Therefore, data mining is one of the most crucial methods of scientific management effective on contact with client increased profitability and client satisfaction. In this paper, using details of the phone cell conversations during two periods (one with no rival and the second one with rival) and the employing details of conversations of the third period for identifying subscribes suffering churn, it has attempted, regarding the pattern of client to predict their churn.