What Lies Beneath—A Survey of Affective Theory Use in Computational Models of Emotion

Abstract

Studying and developing systems that can recognize, express, and “have” emotions is called <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">affective computing</i> . To create a Computational Model of Emotion (CME), one must first identify what kind of system to build, then find emotion theories that match its requirements. The relevant literature is vast. This survey aims to help design CMEs that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">generate emotions</i> —separated into <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">emotion representation</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">elicitation</i> tasks—in computer agents and interfaces. We give an overview of 67 CMEs from different domains, and identify which emotion theories they use and why. To better understand why CMEs use some theories, we also analyze instances where these CMEs use theories to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">express emotion</i> . Lastly we summarize how CMEs generally use each theory. The survey is meant to be a guideline for deciding which affective theories to use for new emotion-generating CME designs.