Abstract

Emotions play an important role in both human intelligence and behavior. Advances in understanding emotions can improve our understanding of human intelligence as well as allow higher fidelity modeling of human behavior. Current research in computational models of emotion is trying to address integration between emotion and traditional cognitive architecture. Different interactions between emotion and cognition have been observed in cognitive science, neuroscience, and psychology. However, only relatively small subsets of the observed phenomena can be explained or replicated by existing computational models. In this paper we identify a small but representative set of phenomena that we believe should be covered by an architectural design to integrate computational emotion and cognition. We discuss a set of computational mechanisms that is sufficient to cover the phenomena discussed and present …

Date

2011

Authors

Jerry Lin, J Blythe, M Zyda

Journal

Under Review