The effects of different alphabets on free text keystroke authentication: A case study on the Korean–English users

Abstract

Keystroke dynamics is one of the representative behavioral biometrics, and it has been consistently recognized as an alternative to physiological biometrics for user authentication to strengthen the level of security. This paper investigates the effects that languages with different alphabets and different familiarity levels have on the free text keystroke authentication performance using Korean–English data collected from 83 Korean participants. In order to exploit the familiarity level, two typing characteristics are measured and tested. Student’s t-test reveals that the participants have higher typing proficiency but lower typing consistency in the language with the more familiar alphabet, i.e., their primary language (Korean). Typing proficiency is found to be a critical factor when only keystroke latencies are utilized during authentication, whereas typing consistency is found to be a critical factor when key sequence information is utilized in addition to keystroke latencies. The experimental results can be applied to build a customized keystroke dynamics-based authentication system, which adaptively determines the authentication method as well as the keystroke size based on a user’s typing characteristics.