Mining Cursor Motions to Find the Gender, Experience, and Feelings of Computer Users

Theory of "embodied cognition" suggests that a variety of mental activities are reflected in states of the body, such as postures, arm movements and facial expressions. The present study investigates the extent to which profiles of computer users -- their gender, feelings, and emotional experience -- can be assessed from movements of computer cursors. In one experiment, participants (N = 372) saw three film clips for two minutes each, rated their feelings afterward, and carried out simple perception tasks three times where our program traced participants' cursor trajectories every 20 milliseconds. We investigated the extent to which the extracted cursor trajectory features could reveal participants' profiles. Results indicated that a small number of trajectory variables were helpful to identify which film participants saw, how they felt while viewing the film, and their gender. We suggest that cursor motions provide rich information for dynamic user profile mining.