Impersonation of Biometric Authentication By Keystroke Rhythm Patterns

Abstract

In this article, we made evaluation on impersonation of biometric authentication by keystroke rhythm patterns. In the experiment, 4 subjects tried to type their own passphrase rhythmically in 4 times, and type others' keystroke rhythm in 3 times after the practice. These collected rhythm patterns were evaluated with Self-Organizing Maps (SOM). As a result, both FAR (False Accept Rate) and FRR (False Reject Rate) are less than 60%. In addition, registering a complex keystroke rhythm pattern, both FAR and FRR are less than 12%. It is possible to contribute to the improvement of reliability by using authentication with key stroke rhythm based on SOM together with existing authentication methods.