The Research And Application Of Telecommunication Precision Marketing Based On Data Mining And Customer Segmentation

With the breaking monopoly and in the new competitive environment at home and abroad.Various telecom operators are making strategic transformation to explore new opportunities in the new technologies and new business brings,to create greater value for the enterprise.Faced with this increasingly fierce market competition,telecom companies urgently need to improve the scientific decision-making capacity and enhance the ability to judge correctly is operating in the market.Telecom companies have accumulated vast amounts of data operations.It is very important transform these data into valuable information for enterprises precision marketing.Therefore,companies need to use data mining techniques to analyze the full range of data,and for the full and effective customer grouping to help companies make the right decisions.Aim at achieving true precision marketing.In this paper,in telecom enterprises based on accurate customer segmentation marketing,the main research and application as follows.First,it introduces the research status based on data mining customer segments in the telecommunications business and in precision marketing applications.And it describes the data associated with customer segmentation mining methods,strategies and algorithms.Especially in several clustering algorithms comparative analysis to conduct a comprehensive in-depth analysis based on the division of K-means clustering algorithm and EM clustering model.Second,data for the telecommunications customer behavior in numerical attribute many features,this paper presents a numerical score to determine the importance of the property value in the property value of cluster analysis.Purpose is to provide the basis for pre-processing of data and provide a reference for assessing the value of the selected property.Third,the characteristics and achieve precision marketing applications for telecommunications issues the goal,to strengthen the assessment of the cluster model.Considering the similarity of the effectiveness of the existing cluster assessment in the mass data set,telecom customer behavior in the time complexity is O(m2),which is difficult to achieve.So this paper presents a cluster of individual quality score.On this basis,to build a comprehensive quality assessment index system clustering,laid the foundation for solving application examples of precision marketing issues.Fourth,in the branch of China Unicom M full investigation,we understand the business and data are available,using CRISP-DM process model,to achieve precision marketing through customer segmentation target.Using K-means algorithm and EM algorithm clustering models were established,and with the evaluation index system based on the establishment of a comprehensive quality model to evaluate and compare.Then select the better performance of the EM customer segmentation model,combined with the practical application of the proposed strategy to achieve precision marketing,to provide a basis for the telecommunications business decisions.