Application of Data Mining for Customer Relationship Management in Securities Industry

• Main Authors: Yung-Ming Chang, 張永明
• Other Authors: Wen-chieh Chuang
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/14867083459512131461

碩士 === 逢甲大學 === 工業工程與系統管理學研究所 === 93 === The development of information technology has change the way business conducted. More and more business transactions are now completed by computers, as a result most companies have accumulated huge amount of transaction data where useful information can be hidden inside. Take the securities industry as an example, the combined customer background data along with their trading history data could conceal useful information about customer transaction behaviors. If somehow these information can be explored, they can be very useful for customer relationship management (CRM) operations. The aim of this thesis is to apply data mining techniques to a real security trading data set. Classification algorithms C5.0 and CART are applied to discover possible relationships between customer background data and transaction data. Three groups – customers that have high total investment money, customers that keep increase their investment money and customer that keep decrease their investment money have been predetermined as exploring targets, several classification rules to identified these target groups are discovered.