Deducing user states of engagement in real time by using a purpose built unobtrusive physiological measurement device: An empirical study and HCI design challenges

Abstract

Human emotion is a psycho-physiological state in most cases not obvious to the subject. Different permutations of emotional constituents sometimes cause similar outward expressions; therefore facial expression methods cannot achieve reliable estimates. Sensing physiological manifestations of hormonal and neural stimulations instigated by emotion and affect is widely accepted as a credible method of detecting psycho-physiological states. A major impediment in interactive environments employing physiological sensing affecting the credibility of measurements is the physical and psychological impairment caused by electrodes and wiring used for the acquisition of signals. In the system described in this paper, the above obstacle has been overcome. Physiological signals acquired via an in-house developed computer mouse and coinciding physiological patterns were investigated as reactions to emotion raising events. A classification algorithm analyzed herein produced a real time allocation model of states of engagement. Experiments have revealed strong correlations between events and respective emotional states.