Mobile Internet User Behavior Based On Association Rulesanalysis

With the rapid development of mobile Internet technology,the intelligent mobile phone,tablet computer,intelligent wearable devices such as mobile intelligent terminal network through marginal,full range into the personal life,work,social and recreational activities.When people use mobile intelligent terminal to access the mobile Internet,a large number of user behavior data can be generated.Therefore,how to make full use of mobile Internet user behavior data and play its potential value,has become the key issues to be solved for mobile Internet companies.However,the traditional lack of user behavior analysis and behavior analysis by means of data mining,mobile users can not distinguish between detailed and accurate demand,found the behavior patterns of users,resulting in the mobile Internet business marketing and user demand does not match,reducing the level of service.Through the research and analysis of the browsing behavior of mobile Internet users,the paper focuses on the user preferences and dynamic preference trends in different contexts.Specifically,the main work is as follows:1.according to the characteristics of the mobile Internet,that is,the user can carry out Internet activities through a variety of mobile terminals anytime and anywhere.This method introduces situational factors,combined with the user information and network access logs for users to browse the Web content classification in order to improve the efficiency of data analysis,access time and access frequency through the Internet on mobile client carries on the statistical analysis,gives the user preference degree calculation method for constructing user preference matrix.Then the association rules algorithm is used to generate the preference relation of user browsing content,so as to explore the user's browsing preference in different contexts.Finally,the method is applied to an example,and its effectiveness is verified by analysis and comparison experiments.2.for association rules is generally in the support confidence framework of mining rules,standards can not be less,dynamic change rule and reaction and the framework of its time,the application of dynamic association rule mining algorithm based on the trend of constructing user preference model,mining user browsing preferences change the law of time,at the same time the rules have higher value.In order to analyze the rationality of the algorithm through an example with the original dynamic association rule mining algorithm(algorithm)were compared,proved that the new algorithm reduces the number of rules to a certain extent,improve the dynamic association rule mining quality.The research shows that by increasing the situational factors,using the dynamic association rules trend degree is built based on user preference mining analysis model can more precise and comprehensive rules,has very important practical significance for Telecom operators.