| Abstract:With the rapid development of various information technologies, information security is facing greater challenges of modern life for a higher information security requirements. Affected by cyber attacks, personal information is often stolen. While the traditional identification methods can only identify passwords instead of the user. The traditional identification methods cannot meet the actual needs already, and the appearance of biometric technology brings hope to solve the problem.Biometric refers to the identification and authentication of humans by their behavioral and physiological characteristics. However, the existing biometric technology just focused on the only physiological characteristics of fingerprint, face, iris, etc. Such recognition technology is not conducive to large-scale promotion, because of the requirement of expensive equipment. While the emergence of keystroke dynamics makes up for the deficiencies in this regard. After the deep research of keystroke dynamics, the paper presents a method based on user identity keystroke behavior.Firstly, this paper focuses on the analysis of the relative content of keystroke dynamics. On this basis, the article focuses on the analysis and research of keystroke characteristics of both static text and dynamic text. Then, the paper describes keystroke data acquisition, data analysis, data processing and feature extraction in detail. Meanwhile, with regard to the proposed feature classification scheme, we conduct the experiment to test the performance of our project and the result shows than the rate of keystroke features of both static and dynamic text areas can reach90%. Meanwhile, the result shows that the continuous identification systems designed in the paper can play a role in the actual environment. Finally, aimed at the feature extraction method proposed by us, the paper carried out a detailed design and implementation, to build a user identification system based on keystroke dynamics. |
PDF Full Text Request