Data Mining About Network User Behaviors

With the rapid development of the telecommunications market and telecommunications services, more and more people obtain information through the Internet, which brought a surge in the number of users. However, when the number of users tends to be constant, ISPs'business development will face a bottleneck. How to change from extensive operation to fine-type operations becomes an important topic for the ISPs. Right in the increasingly competitive Internet economy, only to win the users can gain a competitive advantage.The analysis of the users'behaviors in the network becomes an effective method to support the management and decision and it has been applied to different levels of society. In many important sectors such as retail, finance, telecommunications and insurance industries, a lot of historical data has been accumulated. The policy makers hope to discover the hidden business association rules, customers'classification and other information, and to grasp the characteristics of user behavior, predict the new users through these vast amounts of user behavior data. So they can see the benefits of different businesses, regional differences, the change of business requirements and the proportion of various user types quickly and intuitively in order to cut down the operating costs and provide more human-oriented services.Through some software, we may obtain network packets, and then we will probably analyze the user the behavior through these packets. In this article, we studied the significance of the data mining firstly and the network behavioral analysis. Secondly, we conduct the research about mass data processing, thus design a data processing module. And we carry on the preliminary analysis about user behaviors using the processing result as well as the deeper level analyzing. Finally, we list the next steps and the expandable place.