A linear regression model to detect user emotion for touch input interactive systems

Human emotion plays significant role is affecting our reasoning, learning, cognition and decision making, which in turn may affect usability of interactive systems. Detection of emotion of interactive system users is therefore important, as it can help design for improved user experience. In this work, we propose a model to detect the emotional state of the users of touch screen devices. Although a number of methods were developed to detect human emotion, those are computationally intensive and require setup cost. The model we propose aims to avoid these limitations and make the detection process viable for mobile platforms. We assume three emotional states of a user: positive, negative and neutral. The touch interaction is characterized by a set of seven features, derived from the finger strokes and taps. Our proposed model is a linear combination of these features. The model is developed and validated with empirical data involving 57 participants performing 7 touch input tasks. The validation study demonstrates a high prediction accuracy of 90.47%.