Analysis And Modeling Of Campus User Behavior Based On Accounting System

With the rapid development of Internet technology,especially the development of computer technology,the amount of data in human society has shown an explosive growth trend.How to make full use of the value of data and mine the valuable knowledge and information hidden in massive data It has become the focus of research on Internet companies and universities today.In the field of education,the popularity of computer technology has also promoted the improvement of the level of informationization in colleges and universities.Various types of information management and network systems have been widely used in colleges and universities,and gradually accumulated a large number of student network behavior data.In this context,the use of data Mining technology analyzes the online behavior of campus users,which not only helps campus users to better self-recognition,but also provides effective technical support for network administrators to grasp the use of campus network and the overall online behavior of users.Based on the campus network user access log of Dr.com's billing system of Beijing Jiaotong University,this paper has carried out a series of researches from three aspects:web page classification technology,user behavior analysis modeling,user behavior analysis system design and implementation.The specific research contents are as follows:(1)Research on URL-based web page classification technology.Different from the traditional web content classification method,this paper proposes a URL-based webpage hybrid classification algorithm,which implements URL webpage classification through URL classification library and URL classifier based on N-Gram language model,and finally classifies through WebKB dataset and URL.The library verifies the classification accuracy of the classifier.(2)The user behavior modeling technology based on user attributes is studied.Firstly,data preprocessing is performed on the accounting log,including data cleaning,user identification,session division and session reduction based on web page similarity.Secondly,the topic is classified based on the URL,and the access duration and access frequency factors are introduced,and the weighted interest is obtained according to the weighted interest degree.User interest topic representation;Finally,based on the user's online time attribute,the user behavior model is established,and the accuracy of the model is verified by experimental tests.(3)Design and implement the campus user behavior analysis system.The system backend implements the modules of the user behavior model;the system front end is divided into the campus user end and the campus administrator end,and the campus user end can view the historical access record and the analysis result of the online behavior;the campus administrator can view the whole school.The user's overall access behavior and the ability to detect abnormal behavior.