應用決策樹探討適用於電子行銷市場之區隔基礎

Applying Decision Tree Techniques to Segmentation Bases for E-Marketing

陳亭羽(Ting-Yu Chen) ； 賀千盈(Qian-Ying He)

《管理科學研究》 3卷1期 (2006/06) Pp. 1-25

https://doi.org/10.6873/MSR.200606_3(1).0001

摘要

網際網路的興起帶動了網路購物的潮流，有別於傳統市場通路，網路購物市場有其獨特的特性，在此特性下，是否會使得舊有的市場區隔變數不再適用於網際網路上，加上現今對於網際網路市場區隔變數之研究，較缺乏整體性之探討。本研究以資料探勘中之決策樹技術，探討傳統的區隔變數中，有哪些是仍適用在網路虛擬世界中，或者會有新的區隔變數出現以因應市場結構之改變。另外，將決策樹所得的市場區隔變數與統計方法所得的市場區隔變數，配合統計檢定，檢定各方法孰能較有效得區隔市場。故本研究不但針對網路購物市場，進行市場區隔的步驟，找尋最合適之市場區隔變數，在建立市場區隔後，再探討利用不同的方法所得之市場區隔變數，是否能提供給網路購物經營者不同的資訊，進而在網路市場中，獲取所追求之利益。

關鍵字

市場區隔變數 ； 資料探勘 ； 決策樹 ； 網路購物

並列摘要

The increasing activities on Internet have impacted the conventional market segmentation variables and possibly causing some conventional segmentation variables to become no longer suitable for the Internet environment. Since the current research on Internet market segment variables is not detailed and complete, the aim of this research is to use data mining methods to investigate the adaptability of the conventional segment variables in the virtual world of the Internet and whether there are new segment variables to adapt the change of the market structure. We focus on segmenting the online shopping market, finding the most suitable market segment variables, and establishing new market segmentation through decision tree techniques. In addition, we conduct statistic analysis to determine significant segmentation bases and make a comparison of segmentation results between decision trees and statistic methods. Finally, this study provides the online shopping business operators some marketing aids and hence to generate profit in the Internet market.

並列關鍵字

Market segmentation variable ； data mining ； decision tree ； online shopping