Multi-scale analysis of human behavior patterns of communication based on large scale mobile data

The studies of traditional human behavior dynamics and complex network of human communication are usually based on individuals or the groups. However, urban residents who are separated by urban road network and divided into different parts by spatial distance and geographical constraints, show different characteristics from the individuals and groups. In this paper, we study the dynamical characteristics of human behavior and the features of communication network of residents on different land parcels, with a real large-scale mobile phone data. We find that Functional Urban Areas with different attributes have some influence on behavior patterns of residents. Especially, most of residents are located in the educational land. Furthermore, by contrasting variations of the same indicator in diffrent months. we also find that the number of permanent residents changes the greatest in August in all land parcels. In addition, the dynamical indicators of human behavior and the features of communication networks correspondingly have dramatically changed.