Identity Authentication Model Based On Input Feature

With the rapid development of the Internet,the security of personal information is becoming more and more important.The most frequently used scheme of identity authentication is the password scheme.Traditional text password and graphical password are based on the user's memory,and in case of forgetting them,users often choose short passwords or those which are easily guessed,which increases the risk of invasion on personal information.Therefore,biological identification technologies are getting more and more attention which include chirography,face,fingerprint,etc.In the mobile terminals,when users interact with mobile phones through touch screen,they have their habit of personal behaviors.This kinds of data can be easily collected from mobile phones,which provides the convenience of importing the biological characteristics into identity authentication and effectively improves the security of identity authentication on mobile terminals.In this paper,we introduce the behavioral characteristics in three different ways in the identity authentication on the mobile terminal:(1)The keystroke feature is extracted during the user's input of the text password to authenticate the user(2)the trajectory feature is extracted to authenticate the user during the user's input of graphics password(3)the trajectory feature is extracted during the user's handwriting to authenticate the userIn this paper,we design a data acquisition scheme at first,and then describe the way of preprocessing and extracting features in three ways: text password,graphic password and handwriting.Then we describe the Balanced Random Forest algorithm,Using the principal component analysis to reduce the dimension of the data,and finally the use balanced random forest and principal component analysis to analyze different behavioral characteristics.The experimental results show that the three different behavioral characteristics introduced in the identity authentication of the mobile terminal have good effect on authenticating users.On the data set which dimensions has been reduced by principal component analysis,the execution speed of the algorithm and the effect of behavioral characteristics on identity authentication is ensured.It can be seen that the three different behavioral characteristics introduced in this paper have a good application prospect for improving the security of mobile identity authentication.