Research On Customer Service Information Mining Of Telecom Cpmpany Based On Assosiation Rules

This thesis mainly discussed how to find out intrinsic relationships among telecom value-added services by association rules, which can help telecom companies analyze customer behavior, identify customer feature and make effective strategies for marketing and customer service.At the beginning of reading and analyzing domestic and external documents, this paper summarized the present situation of telecom service management and application of data mining in telecom industry. Then author introduced basic concept and classical algorithm of association rules, general steps and problems of quantitative association rules mining as well as fundamental principles of negative association rules. Next, positive and negative association rules algorithm meliorated by cloud model was presented. There are a mass of data of numerical value type in telecom companies' database. How to reasonably partition domain values is very important in quantitative association rules mining. Traditional method can not get the easy to understand knowledge because it can not reflect, the actual data distribution or the partition is too sharp. In this paper, a new method based on cloud model was introduced. This method can reflect the distribution of data in that domain while keeping the soft boundaries. Therefore, the discovered association rules are also easy to understand. At the same time, this method found out both positive and negative association rules in database with correlation coefficients, and improved the efficiency of algorithm by multiple minimum supports which can restrict the number of short item sets while keep the meaningful long sets. In the end, this paper constructed the system framework of telecom company service mining, discussed the basic model and influence factors of customer behavior, and explored the connection among all kinds of telecom added-value services with positive and negative association rules.