Keystroke dynamics for authentication in smartphone
Jong‐hyuk Roh, Sung-Hun Lee, Soohyung Kim · 2016
When entering a PIN on the touch screen of smartphone, users have their unique patterns, which shows different time interval, strength, position, and usage angel. Using this user's pattern, it is possible to distinguish the imposter who knows the user's PIN, then the PIN verification can be enhanced in the smartphone environment. In this paper, we tested classifiers with five features that can be extracted from smartphone sensors. Also, we collected keystroke data for each user's posture, and compared the characteristics of the posture. We consider the importance of preprocessing and feature combinations.