Data Mining In Web-based Using An Improved GA

In recent years, with the development of data mining technology in database and artificial intelligence technology a new information technology appear, which is extracted the process model from data. At the same time, data mining is a decision support system, which is based on artificial intelligence, machine learning, statistical, and other technologies, highly automated analysis of the original enterprise data, unearthed a potential model, forecast the user's acts to help business decision-makers adjust marketing strategies to reduce risk and make the right decisions. With the rapid development of the World Wide Web, a massive web data links lack of semantic come out, although the AI (Artificial Intelligence), statistics (Statistics), and other mature technology in a variety of specific areas extracted data from the fuzzy implied, the potential useful information and knowledge in the process has been successfully applied, but because of the characteristics of web data, as well as data mining or knowledge discovery itself is an immature, pending further study areas, the study is still a huge challenge. Data Mining on the Web contributes to the study of the comprehensive utilization of network knowledge, and facilitates the enhancement of system security.The study uses a mature intelligent optimization algorithm (GA), and then on the base of the traditional methods of data mining, introduces GA to the data excavation from the Web, process cluster analysis and association rule mining. The analysis of system users clustering provides an important basis for the accurate analysis of the behavior characteristics of user groups, and the adoption of genetic association rule in mining optimizes and enriches the behavior characteristics of the users, makes up incomprehensive analysis constraints due to the shortage of visiting quantity and time, play a forecast optimization role the overall. The main source of the data for processing is Web site topology and user access logs, which is part of the entire system offline. From the large number hits from Web pages of by many users, analysis the behavior characteristics of Web users, and clustering, further analysis each user's personality and behavior features, and get the association rules of user behavior, provide knowledge support adjustment of the decision-making strategy to provide personalized service.