Research And Application Of Telecom Customer Marketing System Based On Data Mining

From ‘Increase broadband speed and lower rates for internet services' to "Mobile number portability?,China is gradually advancing the rigid control of telecom operators.At the same time,the rapid development of the mobile Internet has also intensified the market competition in the telecommunications industry.Under the dual pressure of external policy constraints and internal market competition,telecom operators are facing unfavorable conditions,such as continued decline in revenue and net profit,slowdown in the growth of existing customers.Entering 2019,with the landing of 5G commercial,telecom operators have gradually achieved 5G coverage nationwide.The arrival of the 5G era has brought new opportunities for telecom operators to reverse the current competitive disadvantages.How to do a good job in the promotion of telecommunications services,increase corporate revenue,and do a good job in customer efficient management and quality services to stabilize customer share,is the important development direction for telecom operators in the current.However,in customer management and product marketing,the telecom operators have the disadvantages of extensive pattern,and inadequate integration of data at present,lack of comprehensive guidance in operations,unable to fully reflect the operational advantages brought by the huge data resources of the telecom operators.This article starts from the core needs of telecom operators,applying data mining technology to build a communications customer marketing system.The system mines and analyzes massive telecommunications customer data,researches and proposes value-based telecommunications customer segmentation methods,to optimize customer operation models and marketing resource allocation,and research on telecommunication traffic product recommendation methods on this basis,to ensure improved marketing effectiveness resource investment rate and return on resource delivery,realizing refined operation.The main research work and innovations of this article are as follows:(1)Combining the characteristics of customers' communication behavior,consumption habits,to improve and propose a new telecommunications customer value evaluation model,meanwhile,applying entropy weight method and correlation analysis method to determine the weight of each parameter index in the model,to quantitatively analyze the customer's Value characteristics,analysis results show that,the model has good practical effect.(2)Combining the characteristics of the telecommunications,a Depth-Weighted K-means Clustering algorithm based on Deep Clustering Network is proposed to achieve the segmentation of telecommunications customers.This algorithm improves the problem of poor initialization of the traditional K-means algorithm in the cluster center,and can achieve better clustering performance.Through comparative experiments,it is verified that the clustering algorithm in this paper is superior to other clustering algorithms in the segmentation of telecommunications customers,providing strong information support for operator resource planning,fully embody the practical significance of the system.(3)On the basis of customer value segmentation,in order to further maximize the benefits of marketing resources,the combination of XGBoost algorithm and classifier chain technology is used to implement telecommunications products recommendation.The algorithm fully considers the customer's individual needs,to predict the customer's probability of ordering multiple traffic products,obtains the best recommendation effect,and achieves accurate marketing.Experiments show that,the algorithm used in article is better than other classical algorithms in predicting telecommunications products recommendations.(4)In terms of system design and application,the paper combines the requirements analysis,designs the structure and function of the system,to built the communication customer marketing system.The system provides basic data management,customer operations,product promotion and marketing interface modules for users.Among the modules,customer operation and product promotion functions can realize telecom customer value segmentation and product recommendation prediction based on customer information respectively,Satisfy the user's core operation needs,effectively improve the marketing success rate and resource utilization rate of telecom operators.