在知識經濟時代，資訊的掌握成了企業決勝敗的關鍵，越快掌握正確且充足的資訊，越能處於有力的競爭位置。過去的傳統資料處理方式已跟不上現在的速度，利用資料採礦的方式來做資料的處理及分析，可以有效的發掘出資料隱藏著看不見的重要訊息，將資訊做更大價值的發揮。 本研究以國產汽車維修業為研究對象，針對維修顧客做分析，企圖了解顧客流失的原因。本研究首先將資料採礦與顧客關係管理之概念結合，以建立出研究之架構。其次，將初始的資料透過資料轉換與定義，轉換為使用者需要的結構化的資料，使分析者與系統皆能判讀。利用交互資訊法則來對於欄位做分群，以選出具有特徵項目的欄位。最後透過資料採礦的關聯法則，採用Apriori演算法萃取出對於企業有意義的法則。 本研究顯示出對於企業顧客流失方面上有意義關鏈的法則，同時針對法則的判讀做出解釋並提出解決的建議以及因應的方式，可作為經營決策者制定行銷策略的參考。 In the time of knowledge-based economy, the grasp of information is the key to success. The faster the information can be collected accurately, the more competitive the enterprise would be. The traditional ways of data processing is out of date; however, the method of data mining can deal with useful information effectively. By revealing useful messages behind the raw data, the information collected can be utilized. This research focuses on the study of Chinese Autobile Co. Ltd, in the aspect of car maintenance, analyzing customers’ opinions and seeks to understand the reason of customer churn. First, we will be combining the concept of data mining and customer relationship management in building the main structure of this study. Transforming the row data obtained through data switching and data identification into materials which are good to be read by both analysts and the system is the second step. Using the mutual information to categorize the column and then select out the feature columns. Finally, through the Association Rule in data mining, we will calculate out the most significant rule for this company by Apriori algorithm. The result of this study reveals the key connection between customer churn and the company rule. Meanwhile we also provides the explanation of the rule, suggestions and solutions as well, in order to refer the decision maker in marketing strategy.
