Research And Application Of Massive Users' Mobile And Browsing Behaviors In Mobile Internet

With the development of mobile communication technology and the popularity of smart phones, mobile Internet provides a growing number of users with convenient services, and gradually become an integral part of people's daily life. Hundreds of millions users access to the mobile Internet using their smart mobile devices every day, and produce a large number of user data. The value of data is highly recognized today, how to study and apply massive user data has become the focus of people's attention.In this thesis, we use the mobile and browsing data of massive mobile Internet users, relying on distributed processing framework Hadoop, to study the optimization of user location information, the identification of urban hotspots and other aspects. Besides, we use a universal system framework to implement the visualization of the research results of mobile and browsing behaviors of massive users. User location information is the foundation of the research of mobile behaviors of users. It is significant to get the precise information of users' locations for location prediction,discovery of important locations and so on. This paper presents a method to optimize the location information of users based on mobile Internet data,and we validate the method with real data. Mobile Internet data, which contains mobile and browsing information of users, also provides an opportunity to study the characteristics of urban space. In this thesis, a threshold adaptive method is used to identify hotspots in six major cities of one province in northeast China. The similarities and differences of hotspots characteristics in six cities are concerned. Additionally, we focus on the most important areas of the city - permanent hotspots. We do some explorations of the characteristics of urban space structure. In addition to the study of user mobile and browsing behavior, this thesis also considers how to present the analysis results in an intuitive and vivid way. We designed a universal system framework, and use three examples to describe the implementation of the system.Based on the real mobile Internet data, this thesis analyzes and mining the mobile and browsing behavior of massive mobile Internet users. We obtain a lot of meaningful results. Meanwhile, this thesis also provides a practical implementation of big data visualization, which facilitates the application of the research results.