Design And Implementation Of Campus Network User Behavior Analysis System Based On Data Mining

With the development of China's economy and enhance the level of network information technology,the Internet has gradually penetrated into everyone's daily life.After 20 years of development,the campus network in our colleges and universities get good use and bring great convenience to teachers and students.Meanwhile,as the number of campus network user is increasing,campus network management work is facing more and more problems.Therefore,the campus network user behavior analysis has a very important meaning and value to the campus network construction.In this paper,we use A campus network as example,obtain user access log files,using data mining methods to explore the campus network users' division habit,then provide recommendations for the campus network optimization and build a campus network user behavior analysis system.The main work includes:Obtain user log files of accessing the public network.For user access log files are scattered,not unified format,this paper summarized the files and implemented data cleansing,merging and standardization.After the pre-processing.the field data were understanding.Analyze the target addresses of campus network users.By using the URL and traffic information,this paper introduces the clustering method to achieve the effective division of the campus network users' target addresses,explaining the specific differences in different clusters,and then provide practical data to optimize the export hub for the network.In order to observe the effects of discrete points on the clustering effect,this paper used K-means and K-medoids algorithm,also an improved K-means algorithm combined with the agglomerative method.Analyze the campus network user habits to access addresses.Based on the data of the site's domain name,this paper implements Apriori algorithm,explain the campus network users'habits and preference of accessing sites by getting the association rules,this also helps to understand the surfing situation of the campus network users deeply.This paper starts from the reality of the campus network using,clustering analysis of users on the campus network helps to build public network optimization,association analysis helps to understand the campus network users well,which can help to guide students in universities.Therefore this paper has certain practical significance.