Human computer interaction based intrusion detection

Roman V. Yampolskiy, University of Louisville


In this paper we survey the state of the art in human computer interaction based biometrics which are based on abilities, style, preference, knowledge, or strategy used by people while working with a computer. We examine current research and analyze the types of features used to describe HCI behavior. After comparing accuracy rates for verification of users using different HCI-based biometric approaches we address privacy issues which arise with the use of HCI dependant biometrics. Finally, we present results of our experiments with behavior-based intrusion detection in online game networks based on the strategy utilized by the players.