網際網路的興起與資訊技術的發展，蘊藏在網際網路內的資料持續爆炸性的增加。造成許多使用者浪費了許多時間來瀏覽不必要的資訊。 在以往資料探勘發展至今，已經產生出相當多的分類方法供管理者來分析使用者的偏好與購買意願，真正能適時適性的提供消費者個人化推薦的研究。 因此，本研究嘗試以非監督式類神經網路分類系統中的模糊自適應共振理論網路系統(Fuzzy Adaptive Resonance Theory Network，Fuzzy -ART)，利用使用者在網頁日誌檔的資料，來建構一個可自動對使用者分群，找出使用者喜好網頁的機制。 Due to the development of networking and information technology, the information on the internet is growing by an explosive rate. Consequence people spent an amount of time by reading unnecessary documents on the internet. From the beginning of information mining, different kinds of classification system had been created to determine user`s habits and favorites. In the search ,we try to navigate that it can commend the best product to the customers at the seasonable time. This research shall use the Fuzzy-Art system from Unsupervised Learning Neural Network Classification system, which analyses the travelled and messages from the users and assorts them into different groups our proposed is to advise the user most interested personal web pages.
