| With the quick development and increasing popularity of Internet, network provides people with massive knowledge and becomes the best way of study, education and communication. The feature of network is: including a huge amount of data, diversification of data type, equality, dynamicity, virtualization and so on. It penetrates into the network itself, network service and application. But how to get the knowledge that people need eagerly in the huge network world becomes the focus and hotpot of study. For these reasons, the classifications of network users and user behavior analysis demand to be resolved urgently. Network behavior analysis is an important tool in the network service. It plays an important part in improving the quality of website, enhancing the efficiency of network, ensuring network security, and providing users with the personalized service. It meets the needs of the users and service providers.The thesis proposes the process and system architecture of user behavior analysis based on the concepts of users who surf the internet and user behavior analysis. We consider the efficiency of data acquisition and find an intuitive way to show what a user is interest in. Some critical technologies referred in the architecture are introduced. They are HyperText Transfer Protocol which is analyzed to obtain the basic information of users, packet capture technology, socket programming, data mining, as well as the personalized service technology based on rules and information filtering. It shows how to get the information of the users who surf the net by computer. The users are distinguished by some information that is obtained. We define the formula of interestingness combining with data mining classification algorithm and observe the variation of the interestingness by actual experiment. The values objectively reflect the hobby of these users. At the end of the paper, a number of practical applications of user behavior analysis are stated. |
PDF Full Text Request