Data Mining Application To Customers Relations Management In Security Company

As a medium-sized securities company, Jiang Hai Securities Ltd. has constructed fairly complete computer-based operational system.. To adapt to the increasingly fierce competition of the market, to cater for the shift of management concept from product-centeredness to customer-centeredness, to take full advantage of the abundant valuable information sources yielded from the stock trading system to mine the trading data, and to build and perfect the Customers Relationship Management System of Jiang Hai Securities, it is imperative to enhance the core competitiveness of Jiang Hai Securities.This paper gives a detailed illustration of the theory of the data mining. With the help of the Clementine data mining software of SPSS company, with the basis of trading data of 2007 of Jiang Hai Securities, and with the application of Decision Tree and mining module of Association Rules, the paper mines the trading data and customers information, creates the mining algorithm, realizes subdivision of the categories of the customers, and the prediction of the purchasing intention of the stock investors, and forms the rules to guide customer managers of the company to provide high-quality service for investors.The paper also gives a detailed introduction to the frame and function of Jiang Hai Securities Customer Relationship Management System, and designs the frame of the subsystem of data mining in the Customer Relationship Management System. The subsystem of data mining is mainly divided into three modules: data reading and analyzing, data preprocessing and data mining. Its major function is to perform data preprocessing and data mining on the data prior to data mining, which includes data collecting, simple dimentionality stipulating, cleaning-up of the useless data in the mined data, cleaning-up of the repeated data with Basic Sorted Neighborhood Method, creating customer gender and age attributes according to the ID card information and carrying on the null value predicting and filling on missing attributes with the help of k-means algorithm in Cluster Analysis, and finally integrate and give birth to the basic datasheet needed by data mining.Through the application of the data mining, some achievements have been obtained and applied in the customer management. On the whole, the project provided a guide and reference for the company's following construction of data warehouse and planning of Customer Relationship Management System.