Affective Computing Needs Personalization-And a Character?

The role of emotion and personality is described from several aspects reflecting the current state of the art. First, a general idea of emotion, mood, and personality with their basic theories as well as psychological and computational aspects is discussed. Furthermore, the human interaction with an emotional machine is described. In this scenario, the machine shows signs of emotion and augments the communication with the user by that. Another aspect of affective systems deals with the assessment of human emotion by a computer system. Apart from computer vision, wearable computing, social media, and virtual reality provide data that is used to learn about the emotional state and personality of an interaction partner. This leads to the model of a complete character rather than a system, which just reflects short-term affective reactions to events and perceptions. Part of this character is the personality, but also the personal interaction history, which in turn contributes to habits and environment knowledge which all in all explain the behavior. A further aspect deals with emotion and autonomy. In this scenario, emotion is part of a model, which is used to control an autonomous system. Eventually, application scenarios for emotion-based systems are discussed. Every section deals with emotion as well as with the current state of integrating personality models.