Beware of computers bearing smiles: Modeling the social and cognitive effects of emotion in virtual humans

October 22, 2014 | Aberdeen, MD

Speaker: Jonathan Gratch
Host: ARL Fellows Meeting

Over forty years ago, Herb Simon argued that emotions would be required by any intelligent entity that must act in a dynamic, semipredictable and social world. Nonetheless, the cognitive science revolution largely had comparatively little impact on emotion research. This has change in recent years with greater interest in functional approaches to emotion in psychology and economics, and an explosion of interest in computer science on techniques for recognizing, modeling and exploiting emotions in simulations, decision models, and human-machine interaction. In this interdisciplinary talk, I will broadly overview my research on the role of emotion in understanding the social and cognitive function of emotion for both humans and machines. With regard to cognition, I will review my research on the EMA model of emotion, based on appraisal theory, and illustrate how it predicts the antecedents and consequences of emotion in several decision-making task. With regard to social cognition, I will highlight how expressed emotion fundamentally alters how we make decisions with other decision-makers (both natural and artificial). I will illustrate these points using studies across a variety of domains including medical interviews, economic decision-making and computer games. I will discuss both the theoretical consequences of these findings for human cognition as well as their practical implications for human-computer, computer-mediated and human-robot interaction. Throughout, I will argue the need for an interdisciplinary partnership between the social and computational sciences around to topic of emotion.