Web Cache Strategy On The Weblog Data Mining

WWW is popular for its multimedia transmission and friendly interactivity along with the rapid development of Internet. Although the speed of network has been improved considerably in recent years, the rapid expansion of Internet users, the inherent characteristics of delay in the network and the request/response working mode of WWW still make the network traffic very slow and give no guarantee on the quality of service. Because HTTP has no states, the web server cannot know the user actions and the user requests cannot be predicted. Now the Web server uses the mechanism of cache commonly. It saves the documents that have been visited in the cache by the use of the local character of web access time, thus avoiding of the request to be sent to the local hard disk or other web server, and increasing the respond speed of the server. But existent the cache technology don't consider the pattern of user and the knowledge in the vast web log. With the appearance of the Web data mining, the mined patter can be used for the web cache, and the hit rate of the cache can be increased.Web mining categories and study statuses are introduced at first. Then Web site access procession and hypothesis of user's visit action are analyzed. It is clearly that the problems involved in Web Usage Mining are : how to preprocess the raw web logs to provide an accurate picture of how a site is being used ,how to ensure the user transaction, how to ensure theonly user, how to design effective model of data mining to adapt special applying to put out rules and patterns. Through using the patters and rules, the least recently used algorithm is improved.The web site based on the department of computer scienceand technology, the user access rules are discovered after the cleanup of web logs and data mining, and some problems about the structure of web site are also discovered . At last the cache algorithm is implemented and it is proved that the speed of server response is improved after testing.