Research And Implementation Of Web Data Mining Model Based On Agent

In the ocean of information on Internet, people feel that there are more andmore spread of information and more and more difficulty for searching, So, userswould like to have one "private space", which accords to their own needs to displaythe corresponding information. Based on users' needs and the needs of its owndevelopment, to provide users with personalized services has increasingly become atrend.Web mining and agent technology is the key technologies of Webpersonalization.So,this paper presents a web data mining model based on agent and design apersonalized publication and information release system, which achieved:The personalization of books information retrieval: based on the browsing allthe books, it can adapt quickly to changes in users' interest and in the environment,implement automatically matching between online information and userpersonalized demand model, automatically filter out irrelevant information andenhance retrieving accuracy and efficiency according to the user's interest, love andcharacteristics;The personalization of news view: based on the browsing all the news, itdisplays the most interesting news in users according to the browsing behaviors;The personalization of forum service: based on the browsing and joining theforum, it displays the corresponding forum content according to user's interest;The technology of agent is used widely in the area of personalized informationretrieval. It is used to achieve man-machine mutual learning instead of collecting,filtering, clustering and fusioning so much information. The users only need to posethe requirement or make some options through the system, then the intelligent useragent will transform the users' requirement to the command that the agents candistinguish, search the right one to perform the service and return the result to user.The porcess in transparent to user.