Continuous Authentication Based on User Interaction Behavior

Continuous authentication (CA) is the process which continuously verifying a user based on their on-going interaction with a computer system. In this paper, we propose an adaptive continuous authentication method based on the changes of context, in which providing protection for the user's on-going interaction with computer in different contexts. In order to prevent a situation where an attacker tries to avoid detection by limiting to one input device, we considered both keystroke and mouse usage behavior patterns. In this research, collecting 30 users' data in an uncontrolled environment, extracting the user behavior feature from data by a new feature extraction method, using fusion technology to identify users, and then, according to the recognition result we can judge whether the current user is a real user or not. The experiment result shows that our scheme has a false acceptance rate (FAR) of 0%, a false rejection rate (FRR) of 2.04%, and the authentication time that between 10 seconds and 60 seconds for authentication.