User Affective State Assessment for HCI Systems

HCI field began to see more studies in exploring not only the rational characteristics of human users in making decisions but also the “extra-rational” aspects in interacting with their environment and devices. The first task to exploit one such facet, human affect, is to accurately recognize and assess the affective state in real-time. This paper first serves as a survey of the state-of-the-art in affective state assessment, with focus on computational assessment models. Then a modeling framework developed by the authors based on dynamic Bayesian networks is introduced and compared with other models.