Adaptive Site Web Mining Technology Research

With the swift and violent development of Internet technology and e-commerce, Web is dramatically changing our lives unprecedented. Because more business transactions and servies are carried out through the Web, better services for the need of Web-based applications and understanding the action of customers become the focus of attention today. In order to solve the problems of relationship between customers and providers, adaptive Web sites become to the focus of study at present.This thesis aims to provide a comprehensive research on the principles and algorithm of building adaptive sites. Through analyzing the Web log mining, it proposes a fully new session identification algorithm and trail path complementary algorithm for tree sites containing popup pages, and also provides the improved maximum forward frequent trail path algorithm and object page association algorithm.To design the adaptive site MAWSS, which consists of four modules: data pretreatment, site adaptation, page recommendation and object page association. Among the four modules, the data pretreatment and the site adaptation play a basic and central role respectively.