Internet has become a popular medium for information exchange anddelivery knowledge. Many people get the useful information what theywanted from the Internet and network. Several traditional socialactivities have changed to work in the Internet, like distancelearning and tele-medical system. Traditional buying and sellingactivities also follow the trend. Almost all things will be sold inthe Internet, user will buy the product from the Internet too. Howeverwith the advent of the World Wide Web, online merchant must know whatusers wanted or interests and let user buying something in their site.So recommendation process became an important strategy for themerchants. In this paper we analysis users' behavior and theirinteresting, and then we recommend something to these users. Theanalysis mechanism is based on the correlations among customer,product items, and product features. In this paper we propose analgorithm to classify users into groups and recommend product itemsbased on these classified groups. And the system will help merchantsto make suitable business decision and make personal information tocustomers.
Proceedings of the Seventh International Conference on Distributed Multimedia Systems (DMS 2001)，頁297-304