Personal Trajectory And Social Network Analysis Based On Campus Network

In these years,with the development of the technology,comprehensive WIFI coverage has become more and more common in many places such as universities,enterprises,and large shopping malls.Studies have shown that grasping the time and location information contained in WIFI data can reveal the social relationship of mobile users.However,due to the high complexity of data extraction,there are few studies on the analysis of students’ social behavior through college campus network data.In the past,GPS data was used to compare the shape and geographic distance of two trajectories to measure whether the trajectories were similar.This greatly reduced the similarity accuracy for moving trajectories that could not be detected between APs in WIFI data.In response to the above problems,we conducted in-depth research using the real actual operation data of the wireless campus network of Hunan University,and found that the extraction of semantic trajectories can better deal with the above problems.The specific work is as follows:(1)A new multi-angle algorithm for semantic trajectory similarity(MA-STS)is proposed to estimate the similarity between mobile users.Different from traditional trajectory algorithms,the extraction of semantic trajectories can better reflect human social activity behavior,and use both spatial and temporal dimensions to match the user’s trajectory similarity,thereby improving the accuracy and time complexity of data results.Then,the data is modeled and classified through the machine learning binary classification algorithm,which more accurately measures the social intimacy relationship between users.(2)Infer specific types of social relationships between users,that is,friends,classmates or lovers,by using different intimacy degrees under different time periods and geographic location characteristics.Combined with the decision tree model applied to classification under multi-features,the specific social relationship prediction is carried out and the actual social relationship of students is investigated for verification.Experiments show that the verification effect of this method is ideal.Making full use of the social intimacy reflected by the campus social network can help education administrators better understand the interpersonal relationships of students.(3)Construct the social network graph of individuals in the group based on the MA-STS algorithm,and analyze the importance of individuals in the group through the node importance index in the social network.Apply the method of social network analysis to the mobile behavior analysis of campus wireless network users,and further study their matching trajectory points to calculate the social similarity between different students,taking a class or department as the social relationship between social groups To analyze the social activity of individuals among groups to predict individual mental health problems,it is convenient for education managers to carry out mental health activities according to students,and reduce irreversible tragedies caused by psychological problems.At the same time,the degree of social activity reflects an individual’s ability to disseminate information in social interaction,which can provide a certain reference for counselors to select class cadres.(4)Realize the visualization of social network through python and node.js language,including personal trajectory,social intimacy and heat map display.The use of visualization can more intuitively use the campus network to understand the behavioral trajectories,social intimacy,and crowd density of various campuses in colleges and universities.It can assist school administrators to have a clearer understanding of the behavior of students in the school,so as to achieve different degrees of attention in the management process.Social network analysis technology has always been the key research object of researchers,but there are few researches on obtaining personal trajectories and social analysis using WIFI data in colleges and universities.The trajectory similarity algorithm proposed in this paper provides better accuracy in measuring social intimacy,and applies existing social network analysis techniques to the mobile behavior analysis of campus students,helping educators to better understand students social situation.