| In recent years,with the development of mobile Internet,advertising business has ushered in new changes.In order to attract more users,improve the operating income of the project and reduce the cost of advertising,advertisers hope to be able to analyze the advertising data and project operating data in real time.Therefore,the design of a real-time analysis system that can combine advertising data and project operation data is an indispensable requirement for major advertisers.The real-time operation data analysis system based on advertising is designed and developed to meet the needs of advertisers.The system adopts the solution of embedded point automatic real-time data warehouse to ensure that the data can be analyzed and processed in real time and the results can be found quickly.At the same time,the system still needs to retain the ability of offline analysis to ensure the accuracy of some index results.In addition,the system should also support multi-dimensional,multi model data analysis,and try to reduce the cost of data development.This paper starts from the requirements of the system,mainly describes the system architecture and function design,system implementation and system testing.According to the demand analysis,the system can be divided into two parts,eight modules.The two parts are data center and system server.The data center is responsible for processing data,and the system server is responsible for query,analysis and display.The following are the eight modules of the system.(1)Operation data collection: This module is mainly responsible for collecting the operation data generated by the APP and storing it in the message queue for analysis and processing by the subsequent modules.(2)Automatic real-time data warehouse for buried points: This module is the core module of the system,which proposes a solution of embedded point automation real-time data warehouse.It is mainly responsible for the automatic access of buried point data to the real-time data warehouse to reduce the workload of data developers.(3)Advertising analysis: This module is mainly processing advertising-related data in real time,calculating indicators,and performing visual analysis on advertising data.(4)Behavior analysis: This module is mainly to support analysts to analyze real-time data from the perspective of multi-dimensional and multi-model in a highly liberalized manner.(5)User portrait analysis: This module is mainly responsible for tagging users,analyzing the usage habits of similar users,and further assisting advertisers to make reasonable decisions.(6)Data risk control: This module is mainly responsible for monitoring abnormal changes in indicator data and checking users who have abnormally recharge of IOS.(7)Data management: This module is mainly responsible for managing the data sources queried by the system,the buried events and event attributes of the system,and is convenient for the data center to process data.(8)Authority management: This module is mainly responsible for managing system users,system roles and system permissions,so that the system can be reasonably managed and controlled.The above is the development background of the system and the main content of the paper. |
