標題: 結合決策樹與羅吉斯迴歸以提升顧客保留及探索顧客航空選擇之關鍵因子
Integrating decision tree with logistic regression to enhance customer retention and explore key drivers of passenger's airline selection
作者: 馮欣瑜
Feng, Hsin-Yu
王志軒
Wang, Chih-Hsuan
工業工程與管理系所
關鍵字: 資料採礦;顧客保留;航空選擇;低成本航空;重要度績效分析;data mining;customer retention;airline selection;low-cost carrier;importance-performance analysis
公開日期: 2012
摘要: 隨著全球的國際化,不論目的是為文化交流或是商業交易,海外旅遊的民眾越來越多。而2004年開始,除了既有的全服務航空外,台灣的航空市場開始出現以低價為策略的低成本航空,加上金融海嘯的肆虐,讓原本就已競爭的航空市場,變得更加寸步難行。因此,雖然航空市場越來越旺盛,但也越來越多人來瓜份這塊大餅,可知台灣的航空業變得越來越競爭。面對競爭激烈的市場,以服務為本的航空業,要增加自身的競爭力,唯有想辦法抓住顧客的心,此時就需要建立良好的顧客關係。了解顧客的特質及其所注重的服務,針對這些對症下藥,進行顧客保留及區隔出競爭的航空市場,進一步的擬定策略,才能有效達到企業利益最大化的目標。 本研究本著提升航空業者與顧客之間的關係,使用問卷調查,結合資料探勘的技術,建構分類模型,進行一系列的分析,達到三項提升顧客關係管理的研究目的:(1) 找出影響顧客續搭意願的因子,提供航空業者改善以做到顧客保留;(2) 找出區隔低成本航空及全服務航空的因子,將這兩種型態的航空市場進行區隔;(3) 透過重要度績效分析找出目前航空市場裡,航空業者需改善或保持的服務屬性,以提供航空業者做為策略分析。 在本研究之研究結果發現,(1)使用滿意度資料較能有效分類出續搭意願,因此透過滿意度資料進一步找出影響顧客續搭意願的因子,提供航空業者改善以做到顧客保留;(2)使用顧客服務偏好資料進行分類的結果較好,因此使用顧客服務偏好進一步找出區隔低成本航空及全服務航空的因子,將這兩種型態的航空市場進行區隔。
Owing to the impact of financial Tsunami coupled with global economy recession, the airline industry becomes much more competing than before. Meanwhile, the low-cost carriers appeared in Taiwan since 2004. Accordingly, how to enhance the competitiveness of airlines becomes very important for survival in practice. As we know, airline service is a customer-oriented industry and hence satisfying diverse customers is a key way to achieve the above-mentioned goals. In other words, customer relationship management (CRM) can be carried out to understand what customers desire, satisfy them and result in gaining profits for an airline. In particular, there are three goals to be achieved in this thesis: (1) Finding the critical factors to retain customers, (2) Exploring the key drivers of passengers' airline-selection between full-service carriers and low-cost carriers, and (3) Conducting the important-performance analysis to gain managerial insights and indicate that which service attributes should be first improved. Based on the experimental results, it is concluded that (1) Customer satisfaction can predict customer retention very well, and (2) Customer preference is more effective for classifying airline selection.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT070053326
http://hdl.handle.net/11536/72178
Appears in Collections:Thesis