Dynamics of a computational affective model inspired by Dorner's PSI theory

Abstract

The PSI theory of Dietrich Dorner touches a number of questions, especially about knowledge representation, perception and bounded rationality. However, since it is formulated within psychology, it has relatively little impact on the discussion of emotion modeling within computer science. This paper introduces a computational model for emotion generation and function by formalizing part of Doner's PSI theory. We also borrowed some technical ideas from MicroPSI, one of the concrete implementations of PSI theory by Joscha Bach. Based on this computational model, a number of simulation experiments have been performed and evaluated. The experimental results show that the emotions of agents controlled by our proposed model can emerge from the interaction between the agents and the environment. Then the dynamics of this computational model are studied using Lewis's dynamic theory of emotions. We successfully found hints of phase transitions in the emotional changes, including trigger, self-amplification and self-stabilization phases, as suggested by Lewis. Based on these simulation results, we argue that this computational model is a quite promising approach of modeling both emotion emergence and dynamics. (C) 2011 Elsevier B.V. All rights reserved.