Research Of Telecom Customer Relationship Management System Based On Rough Set

With the deepening of reform of China’s telecommunications industry and the rapiddevelopment of IT, telecommunications management philosophy from the traditional "information-center" into " customer-centric " Therefore, effective use of data mining technologyfor telecommunications companies vast amounts of customer data mining and analysis, andfound a variety of potentially valuable, to fully tap the ability of the customer’s earnings intelecom enterprises to enhance their market competitiveness. Customer relationship management(Customer Relationship Management, CRM) system for its high level of enterprise information,the advantage of the rich data resources and technical experience, telecommunicationsenterprises to improve decision-making efficiency and reduce customer churn rate, the key toenhance the competitiveness of enterprises.Rough set theory as an effective way to deal with an incomplete information, inaccuratedata, with a strong qualitative analysis ability. At present, the rough set theory is applied toTelecom CRM management has attracted the attention of many domestic and foreign experts andscholars. On data mining technology, various data mining methods and algorithms, there is stillinsufficient, pending further study. The specific research for this topic are as follows:First of all, described the research background, purpose, significance, research and methods.In-depth study of the basic theory of the telecom customer relationship management and datamining technology and methods, combined with the theory of Telecom CRM, the use ofassociation rules mining based on rough set technology to solve the problem of Telecom CRM.Second, the rough set data preprocessing, including the handling of redundant data, missingdata filled, discrete data processing. Data quality can seriously affect the data mining algorithmoperating efficiency and application of effects.Finally, this paper studied the predictive model based on rough set customer classification.Based on rough set theory, to obtain data from a telecom CRM system, converted to thecorresponding decision table, then the data items filled, discrete reduction, the establishment ofthe generated decision rules.In this paper, practicality of customer classification prediction model framework validationand analysis. This research work for the telecom enterprise application data mining technology toanalyze consumer behavior and to help improve the level of enterprise CRM, has an importantsignificance in theoretical research and engineering practice.