Research On Customer Experience Management Based On Operation And Maintenance Data And Its Application In Network Optimization

In recent years,the telecommunications market has developed rapidly,and customers have more and more freedom and choice.For operators,customer experience management can help to understand the overall experience of users,and provide users with high-quality,personalized services to enhance market competitiveness.However,the way in which the customer experience is obtained through surveys is too time-consuming and labor-intensive to be widely used in the market.Therefore,how to achieve automatic customer experience management is a very important research topic.The main research content of this paper is to analyze the customer experience through the operator's operation and maintenance data,and to illustrate the feasible application of customer experience management in network optimization.The main work of this paper is as follows:(1)Analyze operation and maintenance data,extract the business processes that users participate in,and express the user behavior information during each business process.This paper proposes a merging method from network connections to user behaviors,which combines the log records of connection-level into the service records of business-level.Each business process corresponds to one user behavior,and the characteristics of one user behavior are described using the attributes of the corresponding business process.(2)Evaluate the customer experience based on the business quality of experience and user behavior patterns.The business quality of experience objectively assesses the quality level of a business process,while the user behavior patterns represent the individualized factors of the customer.In the aspect of experience quality assessment,a multi-level mapping method based on variation coefficient method is used in this paper.In terms of user behavior patterns,this paper proposes two user behavior symbolization methods suitable for different scenarios,and uses PrefixSpan method to perform sequential pattern mining.In addition,the paper also applies and verifies the patterns-mining results through user behavior prediction and user similarity evaluation.(3)Explain the application of customer experience management in network optimization through a visualization system.This paper designs and implements a visualization system.The system can display the quality of experience perceived by users or the service level of base stations based on GIS,and can graphically display statistical information of user behavior semantics,service QoE,and so on.