Customer Relations Management Based On Rough Set

The rough set theory as one kind effective method of processing not integrity and not precise data, greatly has the opportunity without doubt in the data mining domain. The rough set since proposed has obtained the successful application in each domain. It theoretically has the profound significance and the application, is attracting many experts' attention in the world by the huge potential. Specially applies the rough set in the customer relations management, without doubt is one kind of new challenge.This thesis first elaborated the selected topic background, customer relations management as well as rough set theory development and present situation; next, study the domestic and foreign newest achievements. At present, many scholars and organization from different angle of view research customer relations management, But from rough set knowledge discovery and knowledge management angle of view to customer relations management development research extremely few, Moreover speaking of its content, mostly lacks the systematic characteristic, integrity and thorough. The present paper research will further profit from, refines the domestic and foreign correlation research results, and makes the present paper research beginning to base on the home leading level. Through to correlation theories and method summary and contrast analysis, lays the foundation for the present paper thorough research.This thesis has studied the rough set data pretreatment, including redundant data processing, the omission data makes up, the data discretization. Data quality serious influence knowledge discovery algorithm operating efficiency and application effect, therefore this article makes the improvement to the ROUSTIDA algorithm. The inconsistent phenomenon in the reality database is the universal existence, Also studied the reduction method which can process the inconsistent policy-making table directly.This thesis has studied customer classification forecast model which is based on the rough set. Mainly take the rough set theory as the foundation, first gains the data from the CRM system, and convert them into relevant decisiontable. Secondly, we complete and discretize the data in the decisiontable. Thirdly, we reduce the attribute and value. Lastly, we can conclude rules for making decision and establish logic ratiocination system.This thesis validates and analyzes the feasibility of customer classification forecast model. The mining process of customer's knowledge is combined into the system of CRM, constructed intellectual CRM system and realized the automatization between enterprise and customer.