Design Of Identity Recognition System Based On Keystroke Feature

With the rapid development of Internet technology,the importance of personal information security is being increasingly valued.There are many shortcomings in the traditional authentication method that mainly uses user names and passwords,which makes users' personal information at risk of being invaded.At the same time,biometric technology,as a more advanced identification technology,has brought a new solution to this type of security problem.Biometric technology uses the user's physiological or behavioral characteristics to complete the identification and authentication of identity.It has the advantages of uniqueness,irreproducibility,and universality,and has great development prospects.At present,biometric technology is mainly focused on identification based on physiological characteristics such as fingerprints,faces,and irises.Most of these identification methods require additional authentication equipment,which increases costs and affects their popularity.In contrast,keystroke-based identity recognition does not require additional physical equipment to assist,and it can rely on the keyboard to complete identity recognition.In recent years,related research on keystroke recognition has gradually increased.The traditional identity recognition based on keystroke features is mainly based on the user's time feature.Because the types of temporal features are relatively single,the accuracy of recognition is often affected by user's own factors or external environmental factors,lacks stability,and the recognition effect is not satisfactory.In view of the above problems,this article collects the pressure change when the user types the password by adding a pressure sensor in the keyboard.After processing the data,pressure features are extracted from it,which increases the diversity of features,thereby improving the accuracy and stability of the classification model.Then,this paper studies the keystroke recognition of fixed text,and tries the application of random forest algorithm in keystroke recognition.The random forest algorithm is used to complete the construction of keystroke classification model.The experimental results show that the random forest algorithm is used for keystroke recognition and has good recognition rate and stability.Finally,this paper designs an identity recognition system based on keystroke features.It can realize user identification,and has a high recognition rate and good stability.