An Affective Agent for Studying Composite Emotions

Using an affective agent to estimate humans' composite emotions is important for creating believable interactions in human-agent collectives. However, there is a lack of suitable platforms for building large scale datasets on this topic to help researchers improve the accuracy of estimations. In this paper, we design and implement an affective agent, which uses explicit emotion appraisals and a historical group emotion dataset to study a user's hidden emotion compositions.