Mobile Authentication through Touch-Behavior Features

Zhongmin Cai, Chao Shen, Miao Wang, Yunpeng Song, Jialin Wang

Abstract: The increasing use of touchscreen mobile phones to access sensitive and personal data has given rise to the need of secure and usable authentication technique. This paper presents a novel approach to authentication on touchscreen mobile devices. The basic idea was to exploit user interaction data of touchscreen mobile phone to authenticate users based on the way they perform touch operations. A filed study was conducted to design typical touch-operation scenarios and gather users’ data. Behavioral features were extracted to accurately characterize users’ behavior. Diverse classification methods were employed to perform the authentication task. Experiments are included to demonstrate the effectiveness of the proposed approach, which achieves a false-acceptance rate of 4.05%, and a false-rejection rate of 3.27%. This level of accuracy shows that these are indeed identity information in touch behavior that can be used as a mobile authentication mechanism.