Abstract
Biometrics is classified into verification and identification. A lot of researchers of the keystroke dynamics have treated the verification which is used for the user login, based on the assumption that people each type in uniquely characteristic manner. However, its error rate is large compared with other verifications, for instance, by the fingerprint and the retina. In this research, we pay attention to the identification and investigate several characteristics of English text typing on the keystroke dynamics. As an efficient measure, time-interval between press and release of the key (p-r time) is extracted, by which a personal fluctuation of keystroke data is minimized. Experimental results show that the proposed methods on the personal fluctuation are considerably promising.
