Undercover

• DOI: 10.1145/1357054.1357085
• Publication Date: 2008-04-08T12:27:59Z
• Authors: Hirokazu Sasamoto; Nicolas Christin; Eiji Hayashi
• Keywords: 05 social sciences; 0501 psychology and cognitive sciences

A number of recent scams and security attacks (phishing, spyware, fake terminals, ...) hinge on a crook's ability to observe user behavior. In this paper, we describe the design, implementation, evaluation novel class authentication systems that are resilient observation attacks.Our proposal is first rely human simultaneously process multiple sensory inputs authenticate, most attacks. We build prototype based feedback gained through low fidelity tests. conduct within-subjects usability study with 38 participants, which complement analysis.Our results show users can authenticate within times comparable graphical password schemes, relatively error rates, while being considerably better protected against Our design allows us outline principles for observation-resilient systems.