The Research Of Mobile Users' Behavior Based On Apriori

With the development of smart phonestablet PC and 3G service,it's easier and more diverse to get the user's location information.How to model the user's mobile track analyze and predict human behavior had become a hot topic.And it has aroused great interest of many researchers.In conclusion,the current method is mainly divided into two categories.One is modeling the path information collected from GPS terminal,which is mainly used in traditional analysis of intelligent transportation applications.The second is modeling the discrete trajectory of mobile users,which is mainly used in the excavation of human behavior and the prediction of the user's location.The analysis of discrete data can also be divided into two categories.The first is to analyze the characteristics of the movement trajectory.We extract the movement trajectory from the location information.Combined with relevant path information and the habits of mobile users,etc.We can extract useful information that can be applied to the field of advertising push,personalized search and other applications and research.The second is mining the moving rules.We analyze the movement rule of groups and it has high research value in road planning,traffic control,and many other fields.According to the current needs of the user behavior analysis,we analyzed 3G data in different cities and mined individual user's frequent patterns.Then find out the group's frequent patterns from the individual user.The main work of this paper includes:First,we process the data,including the user's filtering and data processing.Second we construct user's trajectory combined with spatial information and time information.Then we analyze the user's trajectory using the Apriori algorithm and find the frequent patterns and the prediction accuracy of the next position.Finally,we find the distribution of frequent patterns of different lengths and in different cities and select the frequent patterns of groups which meet the requirements based on frequent patterns of individual users.