Personalization Web Service Design And Implement

With the rapid development of the Internet, the information available on the Internet increases rapidly. The information diversity on the Internet makes it even harder for users to more effectively find the information which they needed. Usually, we mainly adopt some search engines to search information on the Web. But most search engines return results regardless of user preference. So, most search engines cannot really solve the problems caused by information overload. Users require much better Internet services. It is the direct solution that data mining technology is applied to the process of Web logs mining. By identifying the browsing pattern of users which based browsing activities of users, so we try to improve the site's designing and service for customers, for the website is personalized and become intelligent. The paper puts forward a kind of personalized website model, which would be convenient for users to access by modifying web site and personalization recommendation system.Nowadays, electronic commerce are developing fast and changing traditional dealing mode greatly among global area. Every day, e-commerce web site has huge amount of online dealings, which create huge amount of log files. To the server provider, these data, which can be analyzed to find interesting information about customers, such as their partialities, consume patterns, and even the latent consume patterns, in order to design personal service for different customers. The site wants to attract customers and enhance the competitiveness of the other sites. Therefore, Web mining techniques applied to data mining e-commerce has become a closely watched important area.Personalization recommendation system is built on the basis of web log mining. Web usage mining finds the user access mode through mining the web usage records. All the time, we can improve the performance of web server by analysis and study of the rules in web usage mining records. We can improve the design of the Web, increase the satisfaction of users. All in all, web usage mining has a high value in research.This paper introduces the conception and the basic classification and the produce of Web mining, the method and the main purposes of data mining, which is the basis of personalized Web services. In the process of Web log mining, the pretreatment is the vital work, it affects the entire mining results of quality. Therefore, in this paper, we make a definition for data preprocessing, which pointed out the four main steps of the pretreatment. It includes data purification, users identify conversation, identification and path added. In addition, information technology is also recommended to achieve Web personalization services, key technologies, in order to achieve the needs of individual users, we must use the appropriate Web resources to interested users. This paper has proposed two kinds of recommended methods, these are rule-based and information-filtering. In addition, the article also briefly describes the clustering idea which is applied to the cluster log mining. In the process of the design, with the clustering algorithm we can find the relevant pages, so that the system provides user access to the session and the other page that the user maybe want to access .The paper introduces the basic concepts and fields about personalization and data mining. And it introduces the technologies of web mining firstly. And it put the key stone in describing how to design and realize the model of a personalization system based on web usage mining. And also we test the system which based on the ordinary algorithm .In the last, this paper discusses the real meaning and the existing problem of the system's building, and it discusses what we should do on the next research.In our system, we provide different service for different users. For users, we have promoted the service quality, provided a browsing guide. To the Web site managers, the focus on the user browsing behavior, according to the time of user stays on the page and times of visits and other special data, combined with the user's hobbies, so that users can access pages on their own ways, thereby to achieve personalization.According to result of web usage mining, the system has achieved the following functions: the user must choose way to access the site; for the different user, the website can provide different services; based on the user's information, the system can dynamically adjust the site's page, predict and recommend potential users to access Web links. So, we can build a personalized Web site, allowing users to access more convenient. In addition, in order to give users with different pages, this system is embedded the ASP code, which has realized automatic on dynamic linking function. The paper presents the source program. And the system has been tested by experiments.