Research On Precision Marketiing Of Telecom Value-added Service Products Based On Data Mining Technology

In the telecommunications industry increasingly fierce competition today,value-added service has become the new profit growth point and the major focus fields.With the intuition and experience of marketing staff,it is increasingly difficult for people to find the individual demands.Traditional telecom value-added service marketing strategy has been difficult to adapt to the new form of business development needs.In addition,based on the data mining technique,the precision marketing of telecom value-added services has come into being,which is playing a more and more important role.This paper mainly studies the specific application scenarios of data mining technology in the precision marketing of telecom value-added services,and forms a set of value-added services marketing solutions.Based on the current situation of telecom value-added service marketing,this paper expounds the concept definition of precision marketing and application of precision marketing in marketing management,and introduces the main data mining techniques and methods.At the same time,facing two different types of specific marketing scenarios,this paper designs two different data mining processes and achieves the main steps of data mining based on a large number of data collection and analysis.In the data preprocessing,this paper proposes a new method of data transformation for the distribution characteristics of telecom data.Based on the business objectives,data processing,feature selection,data modeling and so on,which are designed and implemented in order to serve the precision marketing of business in the process of data mining.The main innovations of this paper are as follows:1)For the long tail and multi-peak distribution characteristics of telecom data,this paper proposes a new method of data transformation.Without changing the original data distribution law,based on the area formed by the distribution density function region of the data,the new method stretches and shrinks the data to make data characteristics more prominent.2)In this paper,the classification model is designed and applied to model and market the call reminder service product.In facing of hundreds of input variables,this paper filters the feature subset,and establishes several decision tree models according to the principle of a variety of growth and split of decision trees,and compares and tests the results of different models.Moreover,the appropriate marketing strategy is developed according to the value-added services product characteristic.3)In this paper,the clustering model is designed and applied to model and market the mobile reading service product.Based on the identity of the customer's occupation and behavioral characteristics,this paper sets up subgroups individually and identifies the target group features of the service product through the cross analysis of two dimensional subgroups.Moreover,the appropriate marketing strategy is developed according to the value-added services product characteristic.In the implementation of specific precision marketing,according to the precision marketing theory,precision marketing management platform for telecom value-added service manages and monitors the whole process of marketing.Finally,the effect of precision marketing of telecom value-added service products based on data mining technology has improved definitely and reached the expected target.