The Design And Implemention Of Different Broadband Analysis In User Experience Analysis System

With the mass of signaling data and flow control data joining, the data analysis report system of Anhui mobile, based on the statistical principle,has been unable to fully meet the needs of individual customer. Only in-depth analysis of user behavior, precise identification and market segments,can better offer different levels of customer. At the same time make maximize all parties value,and achieve a win-win situation.This thesis design broadband network analysis subsystem based on the user experience by analysis the signaling data. And the purpose of implementation is to improve the accuracy of different network broadband user identification, provide users with electronic channel preference analysis at the same time. On the one hand provide data support for the business staff to carry out marketing campaigns, on the other hand monitor the number of different broadband users in Anhui mobile customers and enhance the ability of customer insight effectively.First of all, this thesis analyzes the requirements of the system in detail,including functionality and reliability. The overall design of the system based on the demand analysis, and is divided into three modules, which are broadband user identification, user channel preference analysis and application management. The model of different network broadband has a miscellaneous of index data, so combined the two analysis, which is the feature of user’s mobile traffic and gateway contact, to build the recognition model. The electronic channel preference analysis realized by statistic method, use the data of users in the last three months of the electronic channels for business, payment number and user browsing information to find the best marketing channel power. Application management module achieve system login and user management, business functions, offline download function and others. Secondly, this thesis also describes the system of network deployment and software environment, and the precision of user identification reach 54.2%. At last, this thesis makes a brief summary of the design of this system, and prospects its future development from many aspects.