Design And Implementation Of User Behavior Analysis System In Campus Network

With the rapid development of network and information technology,various application systems emerge at the right moment.The number of people using the network increases exponentially.In the process of using network and application system,a lot of data has been accumulated.Through the analysis and mining of massive data,enterprises and individuals can bring some information that is difficult to be obtained by statistical analysis.As the high-density crowd of network application,universities has frequent network visits and numerous campus application systems,which keep the records of the data involving in students and staff on the Internet,their attendance,rewards and punishments,student achievements and the other.Researchers are able to record information of consumption,analyze these data to understand the working status of employees and the learning status of students during working times,provide intuitive support on the data for the comprehensive evaluation of students,and timely find out bad network behaviors and finally adopt targeted education management.For the log data of campus network,this thesis proposes a user behavior analysis system based on the analysis of users' campus behavior attribute and analyzes the campus network data of an independent college as the experimental data to discover the comprehensive characteristics of users' network behavior such as application access preference,Internet access duration and access flow and understand the overall operation law of the campus network.(1)A user individual application preference model was proposed for the analysis of user individual application preference based on interest degree.From the application type,duration,frequency and the traffic of user network access,the interest degree of user application category is grasped in a multidimensional way,and the influence factors are set according to the influence of different dimensions on the analysis data.In addition,the network behavior characteristics of individual users are comprehensively described based on the overall online time and online flow.(2)Based on the user's basic attributes,network behavior attributes and Academic Performance,the user portrait label is constructed to comprehensively analyze the user's individual characteristics.(3)Based on the C4.5 algorithm,this thesis proposes the construction of the user's overall behavior model for the user group's analysis needs of application access,and analyzes the campus network's application access characteristics with the application category as the classification keyword.The influence of network application,network access time and other characteristics on the performance is analyzed with the GPA as the classification label.Combining with the statistical analysis method,we can understand the operation law of the campus network from the perspective of access periods and online duration.Through the analysis and research,this thesis finally depicts the users' individual behavioral portraits tags and group performance clearly through the data Visualization.It concludes the user application preference information list and detects the access features of campus network running applications,and provides reliable and effective analysis results to give managers reference to optimize the campus network and personalized service.The thesis also provides students' network behavioral analysis and precaution for the instructors and counselors in order to give better service to students and the school campus teaching through campus network.