Research On The Applications Of Data Mining In Credit Card Customers Subdivision And Target Marketing

Along with the changing of people's consuming habits, there is a huge growing potential in credit card business in China, but the competition becomes more and more fierce. To hold high-quality customers and upgrade potential high-quality customers are keys for banks to improve competitiveness. Banks can provide higher quality individuation service for customers by getting well known of customers'information and changing this information to knowledge. The technology of data mining refers to finding out some unknown and valuable rules from a large number of credit card transaction data and it is a strong tool for banks to provide individuated service of credit cards.With bank's credit card transaction data and customers'vital statistics data as research object, this paper uses the theory, skills and methods of data mining to acquire knowledge hiding under the information of credit cards and serve for credit card marketing. Its main tasks are following:1. Aimed at characteristics of credit card, it puts forward the index RFMRF for measuring customers'value from the behaviors of both consumption and paying back.2. It introduces the arithmetic of data mining and puts emphases on SOM and FP-tree. Then it subdivides customers into three kinds—high-quality, star and common customers according to their contributions by improved SOM arithmetic.3. It obtains the characters of three kinds of customers by FP-tree and dynamically mines the attributes of customers who are easy to change, so as to search new customers and provide different marketing service for old customers.4. Based on the research result, it brings forward sales and marketing strategy for banks'credit card business and gives a reference for effective sale and marketing of credit cards in future.