EmoTetris for recognition of affective states

One of the recent difficulties researchers face in the area of adaptive computer systems is to get data corresponding to real affective states. In this paper we present a new approach for the induction of affective states in humancomputer interaction. For causing different emotions and obtaining emotion-related data we developed a Tetrisgame (labelled "EmoTetris") and applying it in an experiment. The data collection of two participants while playing EmoTetris took place in a laboratory environment and lasted five weeks, while physiological measurements and subjective data were gathered. Results regarding self-assessments indicate that our induction was successful and recognition rates of emotions based on physiological data are up to 70 percent.