- 學位論文
利用分類技術進行顧客知識建立之研究
Applying Classified Technique to Build Customer Knowledge
摘要
隨著90年代網際網路與資訊科技的快速發展，現代企業經營逐漸走向「顧客導向」的新經營模式，強調為個別顧客量身訂製的個人化及客製化產品，以滿足不同的區隔市場，在不同生命週期之不同需求。Kim(2003)提到透過完善的顧客知識建立可以協助行銷以及第一線人員與顧客互動以增加企業利潤，而顧客知識所包含了顧客輪廓以及顧客區隔等，因此，本研究根據所取得3C流通業之顧客及交易資料庫為研究對象來進行對3C流通業之顧客特徵的了解。 本研究進行程序為，首先利用敘述統計以及迴歸分析方法對3C流通業的顧客特徵作個概括性了解；其次，針對研究目的─找出有價值顧客特徵，以五種情境：顧客忠誠價值高低、消費間隔時間長短、交易頻率多寡、交易金額多寡以及潛在客群，分別利用分類技術：區別分析、決策樹分析及類神經網路來探討高忠誠價值顧客、消費間隔時間短顧客、常來顧客、交易金額多顧客及潛在客群的顧客特徵。 研究結果顯示，透過敘述統計以及迴歸分析方法，企業可以對所擁有顧客有個概括性地了解，以作為進一步分析參考的依據。此外，以分類技術來進行顧客知識的建立是可行的，透過分類技術企業不僅可以建構有價值顧客特徵知識，更可以增加其他方面的顧客知識，使顧客的輪廓更加清楚，以提供知識給行銷人員做行銷規劃或者第一線人員協助其與顧客的互動。
並列摘要
With the fast development of internet network and information technology of the 1990s, modern enterprises managed and moved towards the new type of customer leads, emphasize that makes and melts the products for individualization. Kim (2003) mentions that set up and can help marketing and forefront personnel to interact with customer in order to increase enterprise's profit through perfect customer's knowledge, and customer knowledge include customer profile etc., so, this research focus on the understanding of the characteristic to the customer of 3C industry for the research object according to customers of 3C industry and trade database obtained. This research carries on the procedure, utilize descriptive statistics and regression analysis to understand to the customer of 3C industry characteristic at first; Secondly , - find out the valuable customer's characteristic to studying the purpose, with five kinds of situation: Customer's loyal value level , consuming the size of spacing interval , number of trade frequency , the number of amount of money of trade and potential customer, utilize classified techniques separately: discriminant analysi, decision tree and neural networks to analysis. The result of study shows, utilize descriptive statistics and regression analysis, enterprises can understand having a generality about the customers had , in order to the basis consulted as analysing further. In addition, the setting-up carrying on customer's knowledge with classified techniques is feasible, can not merely build and construct valuable customer's characteristic knowledge through classifying technological enterprises, can increase other customer knowledge of respect even mor , make the customer's profile know further, in order to offer knowledge to marketing personnel to make the interdynamic that marketing planning or forefront personnel helped their and customer.