The Study Of Knowledge Discovery Based On Rough Set And Its Application In CRM

With the development of the information technology and aggravation of market competition, the concept of customer relationship management comes into being. Its key thought is to improve the relation between enterprise and the customers. Enterprise makes a response to every demand in time through managing customer's information resources and meet the particular customer's particular demand by the appropriate channel and in appropriate time. So, enterprise must collect a large number of data and utilize various analytical methods. This way can discover the useful information implied in these datum. Knowledge discovery in database can help enterprise to solve this problem very well, which extracts implicit, unknown, valuable information and knowledge from customer's database or data warehouse according to enterprise's require, turn abundant, unordered data into the mode which describe customer's characteristic. We can say the success of CRM in effective customer knowledge mining.KDD can classify, cluster, association rule mining, etc. This function meets many kinds of demand of enterprise, including: customer group classifying, customer profit ability analyzing, customer satisfaction analyzing, customer loyalty analyzing, customer buy relevance analyzing, etc. However, some model based on KDD technology is more complicated or needs certain field knowledge, which is subjective. As a new kind of mathematic tools, rough set has a lot of advantage compared with the other KDD technology, it needn't quantitative depiction of some characters or attributes given previously. Rough sets ascertains approximation domain of given problem from depiction set directly through indiscernibility relation. Thus we obtained interior rule of this problem. This method is simple and easy to operate. In recent years, the academia has raised to the research upsurge of rough set, research results are more too. However, most research results are about the mining of classification, clustering and association are very few. Accordingly, studying of this paper has important theory meaning...