Wi-Fi Data Based Campus Social Network Analysis

With the rapid development of science and technology, smart phones become the necessities of life, entertainment and communication. Wi-Fi capacity has become an indispensable part of intelligent mobile phone. A Wi-Fi enabled smart phone will send WiFi probe message transmissions(latter called WiFi probes) automatically to surrounding environments to search for available networks. These WiFi probes contain information which can uniquely identify a user. Thus, collecting and observing these data have great significance in studying the mobile users' movement law,the relationships among mobile users' and abnormal behavior monitoring.Based on campus Wi-Fi data set , the paper studies the possible social relationships underlying among smart phone users in a university,uncovers and analyzes the social networks constructed by mobile phone users in the university. On the one hand, the paper trys to uncover social relationships among mobile users from the perspective of semantic trajectory. On this basis, the paper proposes two semantic trajectory similarity estimation named FA-STS (A Flexible Algorithm for Semantic Trajectory Similarity) algorithm and FP-STS (Frequent Pattern based Semantic Trajectory Similarity Algorithm) algorithm. Compared to the traditional trajectory similarity algorithms, the FA-STS algorithm is flexible to control the dimension of locus matching. Meanwhile, the paper demonstrates that this algorithm can be used to infer specific type of social relationships between a pair of mobile users by further studying their matching trajectory points. As to FP-STS algorithm, it is more accurate in extracting user's semantic trajectory model features.Experimental results show that the two algorithms can accurately measure the relationship between users and improve the accuracy of the algorithm. On the other hand, the paper proposes RPC(Resident Population Classification)algorithm, which can be used to extract resident population belonging to different kinds of buildings. Based on the semantic trajectory similarity algorithms proposed above and the Campus Wi-Fi dataset, we uncover and analyze social networks constructed by resident population belonging to different kinds of buildings. Then, this paper studies the characteristics and the forming reason of these networks.