Mining Social Groups in Campus Based on Wireless Detection

With the popularity of wireless networks and the prevalence of personal mobile computing devices, understanding the mobility of users in wireless network is of great significance to the social network analysis. As the usage of computing and communication devices reflects their social connectivity with others, We design a new data collection way which based on wireless detection, and according to the new data collection method, we made a new representation of meeting events and defined meeting time grid, making it more conform to the characteristics of our data. Based on our new definition of meeting event, we clustered user data sets. According to the analysis of clustering results, social network graph was used to analyze campus Wi-Fi users' mobility behavior and linkage correlation, which enabled us to obtain more specific social group features on the basis of the macroscopic distribution of the old cluster results.