Mobile phone communication is an important part of human life. Nowadays a great variety of scientific works are dedicated to the investigation of the calls and cellular networks. Efficient analysis of such networks can be a great source of the information useful for reconstruction and simulation of social phenomena. The main goal of this work was to understand how much information can be obtained from the communication network and build an algorithm, which would help to gather information about users' behavior. We used a calls network to reconstruct the underlying contacts network and to detect the strength of relationship between people in this network. Evaluation results show that the developed algorithm can determine users' classes and relationship between users with high quality. © 2016 IEEE.