資料探勘技術應用於客戶關係管理－以國軍儲蓄獎券為例

Abstract

隨著資訊技術的進步，客戶關係管理與資訊科技的整合，常被企業列為資訊科技投資第一優先，本研究利用資料探勘之技術，探討如何應用在國軍同袍儲蓄會所經辦的軍人儲蓄獎券業務，並且在儲券龐大的歷史交易資料庫中，找出隱藏有用的資訊，同時也希望藉由探勘後的結果，掌握一些儲券的方向，以提供未來營運、相關政策之規劃及主官決策之參考。 經過實際處理7,265,736筆資料後得到有效資料7,155筆，經由資料探勘工具INDEX MINER並使用Apriori法分析，區分現役及退役人員來做分析，發現雖然基於某些主客觀因素，造成購買人數逐年遞減，但仍有某些特定族群對儲券購買仍深具潛力，研究結果期能對儲蓄會未來銷售儲券方向能有所助益。

As a progress in information technology, it is privilege for enterprises to integrate customer relationship management (CRM) and information technology. The R.O.C Military Personnel Saving Administration (PSA) processes the military saving lottery, and this study tries to use the technologies of data mining to find some valuable information in a huge historical transaction database, and this information can be applied to the CRM for PSA to improve its services and handle the trend of customer’s purchase. The mined information could be the source information for PSA’s executive information system (EIS) or the decision information for PSA’s sale policy planning. After processing 7,265,736 the raw data, 7,155 records are use to analyze further. This study tried to use tool (Index Miner) and use Apriori method to analyze. All records are divided into two part, active officer and veteran, to analyze. Although the number of buy saving lottery decrease, there are still some specific groups which are highly interested in buying saving lottery to PSA. This study results wish helpful to PSA’s saving lottery sales in the future.