User Authentication Using Keystroke Dynamics via Crowdsourcing
Andrew Foresi, Reza Samavi · 2019
An increasing number of security breaches in North America are the result of stolen or weak credentials yet many businesses have not adapted their user authentication strategies to account for this vulnerability. This paper presents a preliminary study on a purely statistical keystroke dynamics authentication system that provides an additional layer of security on top of traditional username and password authentication. This form of authentication will reduce the threat of stolen or weak credentials for virtually any system which uses a standard keyboard for authentication. Our model produced an FRR and FAR as low as 2.54% and 0% respectively which is an improvement over other statistical keystroke dynamics authentication models.