Technology Of Identifying WeChat User Usage Based On Network Traffic

With the advance of the information age,smart devices and the applications on then are becoming an indispensable part of people's life.Network traffic,as a carrier for transmitting data during application communication in the network,has received close attention from researcher.Analyzing the network traffic generated by the application can provide important information for activities such as network management,privacy preservation,and user usage identification.WeChat is an instant messaging application that is extremely popular in China and the overseas Chinese users,with functions such as sending text,pictures,voice,video,red packet and fund transfer.However,WeChat has also triggered security issuses such as WeChat red packet gambling,WeChat scams and WeChat phishing attacks while bringing convenience to people's life.This paper conducts an in-depth study on WeChat user usage identification.Firstly,in order to protect user privacy and provide better network services,we propose a technology of WeChat user usage identification based on traffic classification Firstly,the feature information in the network traffic generated by the user usage is analyzed,and the network traffic is segmented into a plurality of subsequences representing the user usage in a fixed time duration,and traffic features is extracted from every subsequence,and then the traffic classification model is trained by using the Random Forest algorithm.The results of the evaluation experiments on the real dataset show that the average F1-score of identifying user usage is mote than 86%,and the F1-score of WeChat texts and images can achieve more than 96%,while the F1-score of the staged usage red packet and fund transfer is up to 83%.Secondly,in order to analyze the illegal behaviors such as WeChat gambling,WeChat scam and WeChat phishing attack,we proposes a technology of identifying staged usage based on traffic classification for WeChat users.Firstly,the specificity of the staged usage and the characteristic information in the network traffic generated in each stage are analyzed,and then the network traffic is divided into different subsequences at the packet time interval.In every subsequence,the statistical feature information is extracted according to the traffic characteristic information of different stages,and the traffic classification model is trained by using the Random Forest algorithm.The experimental results show that our method can identify all stages of the staged usage,and the average FI-score rate is more than 96%.In summary,the proposed technology of identifying WeChat user usage and staged usage is feasible,which can be applied to network management,privacy protection and user usage identification.Therefore,this paper has certain theory and use value.