Multidimensional views on mobile call network

Abstract

Recent studies on social network have spurred significant interests in human behaviors. Nowadays, various kinds of interpersonal human interactions, from mobile calls to emails, provide particular avenues to explore the inherent properties of communication patterns. In this article, we present a comprehensive study on a massive anonymous call records obtained from a major mobile service operator. The important difference laid in our work and previous mainly topological analyses is that we report on multiple aspects of the dataset. By investigating the calls of the users, we find out that most calls tend to last within one minute. Call duration between two females is much longer than that of two males. But calls of males generally involve more stations than that of female, indicating a larger mobile range of the males. We also observed that people tend to communicate more with each other when they share similar characters. Besides, the network is well-connected and robust to random attack. We also demonstrate that the close-knit sub-groups with little discrepancy in the characteristics of its involved users usually evoke more calls. Another interesting discovery is that call behaviors among people between workdays and weekends is obviously distinct. Generally speaking, the goal that we research on call network through multidimensional analyses is to uncover the intricate patterns of human communications and put up reasonable insights into future service intelligence.

Keywords

social network / multidimensional views / statistic properties / dynamic analysis / call graphs / communication pattern