I Know It's You: Touch Behavioral Characteristics Recognition on Smartphone Based on Pattern Password

In recent years, pattern password has been widely used for user authentication on smartphones and other mobile devices in addition to the traditional password protection approach. However, pattern password authentication mechanism is incapable of protecting users from losses when a user's login credential information is stolen. We propose an identity verification scheme based on user’s touching behaviors when inputting a pattern password on the smartphone screen. By exploiting the biometrical features, such as position, pressure, size, and time when a user inputs a pattern password to a smartphone, the proposed user verification mechanism can validate whether the user is the true owner of the smartphone. We adopted fuzzy logic, artificial neural network, and support vector machine, to build classifiers, using the behavioral data collected from 10 users. The experimental results show that all the three algorithms have significant recognition capacity, and the fuzzy logic algorithm is the best one with its false acceptance rate and false rejection rate as 4.7% and 4.468% respectively.