Controlling Adaptation in Affective Serious Games

Managing the adaptation in affective serious games is established mainly on the basis of player emotions. Emotions can be recognized by analyzing physiological data and facial expressions of the individual player. While physiological data provide information about the measurement time, electrodermal activity, pulse, temperature, neural activity, etc., facial expressions are a source of data about the emotion extracted from images. The paper presents a general workflow of game adaptation control in affective games applying emotional state, game outcomes and efficiency, and playing style. It provides preliminary results received by using clustering algorithms for recognizing emotions from physiological data and by applying a convolutional neural network using images to recognize emotions from facial expressions.