The Design And Implementation Of Statistical Analytics And Management Platform For Mobile Applications

In recent years,the rapid development of mobile Internet promotes the popularity of mobile phones,tablet,smart hardware and other mobile terminals,and also promotes the popularity of mobile applications,which provides more in-depth services for mobile Internet users.Now competition among mobile applications has become intense.Mobile application developers,operations managers and other stakeholders face with challenges of lean operations,precision marketing promotion,and user–segment classification,which depend on user data.To achieve these objectives,user data collection and analyzing is indispensable.However,due to the variety of mobile applications distribution channels and multiple environment of users,scarce network and computing resources,people who need to collect and analyze user data have to cope with some problems,such as: how to design collection content and structure to avoid unreasonable consumption of network resources and computing resources on mobile devices,how to handle high concurrent performance problems on the server,how to allocate storage and computing resources on the server side,and how to balance the data validation and user privacy,and so on.In this thesis,the key problems in user data statistical analytics of mobile applications are studied.A statistical analytics and management platform of mobile applications has been developed using the technologies of Node.js,MongoDB,Hive and REST.It implements user data collection from mobile devices,basic data analytics,customized business data analytics,and data visualization capabilitiesResearch works of this thesis are as follows:1.Requirements analysis.The thesis describes the functional and non-functional requirements of statistical analytics and management platform,and build use case models of user data collection from mobile devices,user data analytics and customized business data analytics.2.Architecture design and key technologies research.The architecture of platform is designed from three aspects: module division,data view and deployment view.Three key technologies are studied,including basic data rapid analysis,business data customization,and business data visualization configuration.A technical solution of high concurrent processing is proposed,using Node.js and MongoDB.3.Detail design and implementation.Mobile data collection SDK is designed and implemented to optimize the data transfer between mobile devices and data collection service,which provides a high degree of concurrency and fast access of non-blocking I/O.The basic data analytics is designed and implemented to support three-dimensional analytics and customer loyalty analytics,while the customized business data analytics is implemented with the capabilities of big data analytics.4.Testing and deployment.The statistical analytics and management platform for mobile applications has passed the function test and performance test.It is finally deployed on the cloud platform of Wonders Information Co.LTD.The result of testing and deploying verifies the correctness of platform function realization,and shows that the platform can meet the performance requirements.The platform runs steadily since it gone live several months ago,and supports mobile business operation as expected.