A Preliminary Study of Finger Area and Keystroke Dynamics Using Numeric Keypad With Random Numbers on Android Phones

The paper investigates the effectiveness of finger areas and keystrokes dynamics data as the authentication factors using random numeric keypad of touchscreen smartphones. The users' finger areas and keystrokes dynamics data are collected via android application. Preliminary experiment was conducted, involving 30 subjects typing a predefined 10-digit numbers for ten times continuously. The features used in the analysis are gender of the subjects, dwell times, flight times and finger areas. Random forest classifier is used and the weighted results were true positive rate of 80.7%, false positive rate of 0.7%, precision of 81.3% and recall of 80.7%.