Discovering user access patterns on the World Wide Web

摘要: The World Wide Web provides its users with almost unlimited access to documents on the Internet. use of intelligent agents is suggested assist locate related their interests instead browsing via primitive search engines. A number key components in such systems are identified and a system architecture proposed. In particular, learning agent designed along underlying algorithms for discovery areas interest from user logs. discovered topics can be used improve efficiency information retrieval by prefetching storing then document database system. prototype has also been implemented illustrate various concepts. Experiments performed which show that area fact distributed as