Agents That Talk Back (Sometimes): Filter Programs for Affective Communication

This paper introduces a model of interaction between users and animated agents as well as inter-agent interaction that supports basic features of affective conversation. As essential requirements for animated agents’ capability to engage in and exhibit affective communication we motivate reasoning about emotion and emotion expression, personality, and social role awareness. The main contribution of our paper is the discussion of so-called ‘filter programs’ that may qualify an agent’s expression of its emotional state by its personality and the social context. All of the mental concepts that determine emotion expression, such as emotional state, personality, standards, and attitudes, have associated intensities for fine-tuning the agent’s reactions in useradapted environments.