Keystroke Dynamics for Continuous Authentication
Ananya Ananya, Saurabh Singh · 2018
Authentication systems verify the genuineness of the user who is accessing certain services. The verification process by these systems is done when user logs into the system. These services remain available until user logs off from the session. In this article we discuss a different authentication system based on keystroke dynamics that monitors the user continuously during the session. This system ensures that, the user accessing the services during the entire session is same who has started the session. Previous continuous authentication systems required preregistration in order to create training template. In the proposed work we devise a method that does not require any preregistration. Here a 26-by-26 (aa, ab...zy, zz) matrix is used to collect user's data during the session and the same is used to verify her. To measure the performance of the proposed system, we propose new metrics Strokes to False Reject (SFR) and Stroke to False Accept (SFA).