Mobile Internet Users Behaviour Analysing Based On Cloud Computing

Nowadays, Internet has been a part of people’s daily lives, which makes people get rid of the original lack of information era, into the information explosion era. Be faced with such a large amount of data, no one has enough energy to study all of these information. This makes people feel helpless. To solve this problem, we need to analyze the user’s behavior and the user’s interest. Then we will be able to recommend interesting information to users. Both the users and the information providers can benefit from this way. It helps users save the time of finding useful information and helps information providers reduce the cost of information push.The thesis introduces the research background, significance and research status first. Then it introduces the characteristics of mobile Internet user behavior, the content and methods of mobile Internet user behavior analysis and the application of data mining in mobile Internet user behavior analysis. Subsequently, Hadoop technology, MapReduce programming framework and Hadoop distributed file system are listed to solve the problem of massive data processing difficulties. And then, the thesis briefly describes the data collection process and the pretreatment process. After that, the pattern of users accessing the server, spatial and temporal dynamics of mobile Internet services, collaborative filtering algorithm and user interest recommendation system are described in detail. In order to explain the pattern of users accessing the server, I analyze the number of users accessing the server and the number of servers user accessing. According to the analysis result, I classify the users and the serves into some groups. Spatial and temporal dynamics of mobile Internet services is about the relationship between flow traffic and users spatial and temporal dynamics. Finally, the thesis introduces the result of recommendation system experiments using mahout.