本研究利用資料庫知識發現流程構建一忠誠旅客分析型顧客關係管理系統，本系統主要由兩部份所組成，首先透過分析網路程序法構建旅客忠誠度評估模式，再應用C5.0決策樹分析探勘不同類型忠誠旅客。本研究以某國際航空公司為例，研究結果顯示會員類型、廣播服務、航空公司外觀形象、飛機準時起降、訂位人員態度能區別真正忠誠旅客和態度忠誠旅客。最後，本研究結果不僅可提供航空業者發展客製化旅客保留管理策略，並可作為建構旅客資訊管理系統之參考。

Analytical customer relationship management (CRM) is critical to airlines for retaining loyal passengers. This study develops an analytical CRM system for managing different loyalty passengers using the knowledge discovery in database (KDD) process. Primarily two parts are composed the KDD process, including a weighted- evaluated loyalty model and a data mining technique-C5.0 decision tree. In addition, a case study of Taiwanese airline is applied to examine the theoretical process. Research results found that the member category and satisfactions with clear announcements during flight, airline physical image, reservation staff attitude, and airline punctuality represent differences between true-loyalty and attitudinal-loyalty passengers. Finally, the results of this study are not only useful for other Taiwanese airlines but can also be used to expand management information systems (MIS) to CRM for the airline industry.