| 摘要: | 現在，對於公司，顧客關係管理成為了一門不可忽視的學問，在決策的制定上，它更成為了一樣重要依據，而顧客管理的核心為準確區分出不同類型的顧客，並依照其特性給予刺激，無法留下的顧客是可以避免的，潛在的顧客是藉由適當地刺激來活化的。現在，在顧客區分上，最廣為使用的方法為靜態的顧客區分分析方法，因為藉由已得到的數據進行分析是最為方便的，但是，隨著大家對於顧客關係管理的重視日漸增加，時間變成了必須被考慮的因素，顧客的消費行為本就為動態的，隨著時間改變，顧客的消費行為會有所不同，因此，以時間序列的方式進行顧客分析為現在公司的首選方案。
在台灣，航空公司各家爭鳴，面對市場的變化以及同業的競爭，顧客關係的維持是不可忽視的，忠實型顧客的培養變得格外重要，為了建立長期的顧客關係，完好的顧客關係管理是首要條件。在本研究中，使用了在顧客關係管理領域常用的RFM模型為基底，並應應行業加入了新的指標，使其在制定顧客價值上效果更好。此外，顧客消費行為的分析上，我們使用了時間序列的方式進行分析，以此藉由了解顧客的消費趨勢來更加準確地判斷顧客的類型，進而給予其相對應的方案，藉此維繫良好的顧客關係。
;To modern companies, Customer relationship management is exactly an indispensable part of the company. In making the decision, it also becomes a vital criterion. By the core of Customer relationship management, we precisely classify different types of customers and give suitable stimulations corresponding to their characteristics. Hence, those who are not left are relatively avoidable. Potential consumers are capable of activating through some proper stimuli. Nowadays, when differentiating customers, we almost believe that the commonest method is the static partition method. It is the most convenient way that we only analyze some given data. However, everyone attaches importance to customer relationship management progressively. Time must be considered as a critical factor. As time goes on, consumer behavior is going to change constantly, since customer behavior is indeed dynamic. As a result, it is exactly a primary choice that we analyze customers based on time series data mining.
In Taiwan, many airline companies are facing great variations in the market and tremendous competition from others. Maintaining a positive relationship with customers is not ignored, cultivating customers’ loyalties becomes in fact still more important as well. In order to establish a long-term relationship, comprehensive customer management is a primary principle. In this study, the REM model used commonly in the field of customer relationship management, and the new indicators, applying to airline industries, better and more efficiently evaluate the value of customers. Furthermore, we use time series data mining regarding consumer behavior to show trends of customer behavior. In addition, we come up with an optimal strategy to keep connecting a favorable relationship in advance. |
