Research On Hospital Dynamic Identity Authentication System Based On Key-mouse Behavior

With the rapid development of Internet technology,the process of hospital information is gradually accelerating.Nowadays,medical work relies on the assistance of information system to a great extent.Once the hospital information system is subject to security threats,it will affect the development of various medical work and infringe on patients’ privacy.To construct a safe and efficient information system is the most important problem in hospital information construction.Identity Authentication is the primary measure to ensure information security.Traditional user identity authentication methods include user name and password authentication,certificate authentication,biometric authentication,SMS authentication,and third-party login authentication.The above authentication method does not authenticate users after they pass the first authentication.Therefore,the security of user information cannot be guaranteed continuously.This paper has fully studied and analyzed the latest research progress of user identity authentication at home and abroad,and proposed a user dynamic identity authentication method according to the keystroke behavior characteristics and mouse behavior characteristics of medical workers.The specific research content of this paper is as follows:(1)Based on convolutional neural network and cyclic neural network,a user mouse behavior authentication method based on CGU model is constructed.In this method,the features of mouse drag behavior,mouse click behavior and mouse movement behavior are fused in the way of RGB image three-channel,and then the CGU model is used for experimental analysis.(2)Construct a user keystroke behavior authentication method based on OLG model.The method divides the keyboard keystroke behavior characteristics into UU,UD,DU,DD four time interval characteristics and takes them as the input data of OLG model.Then the experimental analysis is carried out to further excavate the hidden keystroke behavior of the keystroke sequence and improve the accuracy of user identity authentication.(3)The decision fusion method is adopted based on the classification results of the user’s keystroke behavior and the classification results based on the user’s mouse behavior to comprehensively analyze whether the user is legitimate and increase the reliability of user identity authentication.(4)This paper studies and implements a dynamic identity authentication system based on user key-and-mouse behavior.It includes user keystroke behavior authentication submodule,user mouse behavior authentication submodule,user keymouse mixed behavior authentication submodule and trust model.You can judge which submodule to call according to the type of user behavior,and then output the result,and finally get the final trust value through the trust model.It ensures efficient completion of user identity authentication and saves resource consumption.