A framework for modeling emotion in synthetic agent

Emotion is an important part of human intelligence. They can influence human behaviors and guide human to adapt to the environment. Integrating emotion into agent can enhance the authenticity and creditability of agent´s behavior. This paper classifies emotion as cognitive emotion and reflex emotion. The generating mechanisms of these two emotions are analyzed. In order to model emotion in agent, a framework is proposed. In the framework, a production system is used to model cognitive emotion based on appraisal theory, and the reflex emotion is generated according to conditioned emotional response theory. At last, an experiment is given to evaluate the proposed framework.