An Analysis Method Of Abnormal Behavior Of Network Users Based On Behavior Sequence

With the development of Internet,various kinds of network applications emerge in an endless stream,providing convenience for users to access the Internet.The Internet has become an indispensable part of people's lives.But at the same time,it apperars the abnormal traffic behavior of various network users.There are many ways to discover and solve the abnormal traffic behavior of users effectively,but the results obtained are different from perspectives.The main problems of user traffic behavior analysis are the lack of systematic definition of network user traffic behavior,and the high degree of description of network user traffic behavior is difficult to describe.In order to solve those problems,the thesis based on the similarity of users' network behaviors to mine user behavior habits and identify abnormal behaviors of users,which has strong theoretical significance and practical application value.The main contents include:(1)The definition of network user behavior sequence: the current academic network behavior sequence is not defined,so the thesis define the user behavior sequence in the network according to the current network data,and each user grasping network data into a user behavior sequence.(2)Network user behavior similarity and correlation: for all the users in the network,users are screened under relevant rules considering the factors that are few user behaviors or user factors.After screening,each user needs to use similarity algorithm to solve user behavior similarity and finds out the potential relationship between users and internal relations.(3)Network users' abnormal behavior analysis under the network environment: Based on the similarity and correlation between users' behavior sequences,we can find out the possible abnormal behaviors in user's network behavior by comparing similarities.(4)The enterprise network user behavior Mining: in the premise of behavior similarity between the sequence and the user behavior,this thesis can get the same similarity between users,and use text clustering method to mine user network behavior,then achieve the classification of enterprise network users for the same user network behavior habits summary model.