Emotions and Mood States: Modeling, Elicitation, and Classification

Common experience suggests that emotions cannot be reduced to single word definitions. Researchers exploring the subjective experience of emotions have pointed out that emotions are highly intercorrelated both within and between the subjects reporting them [61, 62]. For example, subjects rarely describe feeling a specific positive emotion without also claiming to feel other positive emotions [63]. This high variability in expression and definition of emotions implies that the development of an automatic emotion recognition system is a very challenging task. In this chapter, three crucial issues are addressed: modeling, elicitation and classification of emotions. Moreover, the role of the Autonomic Nervous System (ANS) patterns is emphasized along with the related nonlinear dynamics.