隨著資訊科技的發展,企業內部資料量遽增,其主要原因是這兩年網路技術及創新服務的興起,如何善加利用所收集到的顧客資料,去整合且發現潛藏在裡面有用的資訊。透過資料探勘技術,萃取轉換成有用的資訊做有效利用,提供給不同部門組織做決策之參考,為市場行銷帶來關鍵性且重要決策的依據,此亦成為企業經營及價值創造上非常重要的課題。 本研究以大型零售量販店為研究對象,針對顧客歷史交易資料進行顧客區隔分析及關聯分析。首先,第一階段採用RFM模型為基礎,利用此三個變數進行顧客分群;第二階段我們將運用分群結果,檢視各群之人口基本特性,依分析結果提出顧客價值的區隔建議,並在將來在面對顧客或是新顧客時,都能輕易辨識高價值顧客,並針對不同群組顧客給予不同的行銷策略,及對各價值群組推適合之產品組合。
With the advancing of information technology and intense competitions, data mining technique has been used generally in industries with enormous or complicated data. From the large detection information to discover some neglected or hidden important rules of data for reference as business operating strategy. This study was applied data mining techinques to analyze customers’ profiles and transaction data. There are two main stages in this data analysis. In the first step, we use RFM model to identify appropriate cluster number for the studied customer data. Then, it proved that clustering techinques and association approaches was useful in setting marketing segments and consumers’ behavior. An association rule mining approach was implemented to provide consumers’ behavior characteristics and their product mixes. The purpose of this thesis is to provide a complete data analysis process. And the customers is divided into high, middle and low customers of value groups. We expected that the marketing staff of retailing business can use the proposed procedure as soon as they need to identify the important customer. Finally, this research develops different marketing strategies for every customer group of different customer value. Results in this research can provide a valuable reference about customer relation management for managers of retail industry.