資料探勘在投信業之應用–以共同基金為例

Data Mining Application an Investment Trust & Consulting – A Case of Foundation

楊宗明(Tsung-Ming Yang)

指導教授 ： 邱昭彰 博士

元智大學/資訊學院/資訊管理學系/碩士(2005年)

摘要

行銷策略之擬定，必須根據市場資訊獲得，運用統計分析方法，產生市場資訊。本研究以投信業者為研究對象，同時運用資料探勘的技術，探討其客戶交易資料與基金間的特徵與行為。希望能進一步了解市場資訊，透過探勘的技術以利行銷決策的制定。
面對激烈的商業市場，企業在擬定市場策略時，對所需的行銷市場須加以區隔與分析，並找出各個區隔市場中顧客的消費行為與特徵，以確認其目標市場與目標顧客群，假以適當的行銷策略，才能有效的運用資源，取得競爭優勢。過去企業重視的是市場佔有率，現在更重要的是顧客佔有率，尤其是顧客價值。
從本研究上發現，Kohonen(SOM)分群模式在本研究案例中能獲得較佳的顧客區隔，並可以找到最佳的群集數，以三群集為顧客區隔。在各個區隔市場中我們可以清楚發現顧客的消費行為與特徵。而透過 Kohonen(SOM)分群模式更可以清楚的了解，哪些顧客是投信業最具有價值及潛在性的客戶。我們也發現當顧客與企業往來的次數愈高，則為企業有利潤之顧客；相對也能帶來更多的收入，成為企業最有價值的客戶。以顧客為導向的生產與服務絕對是未來產業發展的趨勢。
未來如何運用資料挖掘的技術，來提高企業的競爭優勢及服務品質來保留現有的客戶，同時還可以選擇更多適合客戶的產品，以協助企業制定，具備競爭優勢的策略行為，是我們未來努力的方向。

關鍵字

投信 ； 基金 ； 資料探勘 ； 群聚分析

並列摘要

The marketing strategy must be following market trend, and use statistical analysis to come out market information. The target of this research is investment trust and consulting, besides we use the data mining to analyze customer’s dealing details and foundation’s character and behaviors, then to get further marketing information to make marketing strategy.
Facing a competitive market, while the businesses make marketing strategy it has to analyze and segment the potion of market, try to find out the consumer behaviors and characters, then to make sure target customers and market. The business should use a suitable marketing strategy to get advantage. In the past, the businesses focused on marketing share, nowadays not only focus on customers share but also emphasize on value of customers.
According to this research, we found Kohonen(SOM) clustering can take better segment of customers, then to divide three groups which can find out the consumers behaviors and characters. Through Kohonen(SOM) clustering we can understand which customers is the more valuable and potential for investment trust and consulting. Besides, we found more deals between customers and business which is not only profitable customers for business; also bring more revenue for business. Absolutely the guideline of customer for produce and service is trend in the future.
However how to use data mining to enhance competition of advantage and service, to keep exiting customers also choosing more competitive products for customers. In terms of making a competitive strategy which is our future goal.

並列關鍵字

Investment ； Foundation ； Data Mining ； Clustering

參考文獻

[10] 陳俊任，「應用資料挖掘於顧客關係管理之研究 - 以化妝品業為例」，元智大學，碩士論文 2001。

[25] 簡榮廷，「多尺度資料的二階段分群演算法」，元智大學，碩士論文 2003。

[4] Chengyi, S., Xueli, Y., and Xinfang, F., “Adaptive Clustering of Stock Prices Data Using Cascaded Competitive Learning Neural Networks,” Man and Cybernetics, IEEE International Conference on, Vol. 3, No. 1, pp. 2359-2363, 1996.

[5] Daskalaki, S., Kopanas, I., Goudara, M., and Avouris, N., “Data Mining for Decision Support on Customer Insolvency in Telecommunication Business,” European Journal of Operational Research, Vol. 145, Issue. 2, pp. 239-255, 2003.

[6] Duboff, R. S., “Marketing to Maximize Profitability,” Journal Business Strategy, Vol. 13, No. 6, pp. 10-13, 1992.