由於零售商面臨動態的競爭環境再加上消費者對產品需求多樣化的消費模式。面對如此的競爭環境,企業常利用顧客關係管理(customer relationship management,CRM)與資料庫知識發掘(knowledge development in database,KDD)來掌握最佳競爭優勢。本研究的主要目的是以某零售業的顧客基本資料與歷史交易資料來建構潛在客戶購買預測模式,使用分類迴歸樹(Classification and Regression Trees,CART)找出潛在客戶及流失客戶並做比較及瞭解潛在客戶的購買行為。本研究以366,349的會員基本資料與1,217,142筆交易資料為樣本。研究結果發現所建立的模型有相當高的預測正確率而可以比較並瞭解出潛在客戶的購買行為。
Retailers now face a dynamic and competitive retail environment and diverse demands of customers. Accordingly, customer relationship management(CRM)and knowledge development in database(KDD)are used by retailers as means of standing out from the competition. This research is aimed at building a screening model which comes from customer profiles and order history, and Classification and Regression Tree (CART) was used to identify the potential customers. To evaluate the performance of the proposed method, the dataset provided by one of the most famous retailing company in Taiwan is used for analysis. The result shows that the constructed model can identify the potential customer successfully.