Clustering ECG Time Series for the Quantification of Physiological Reactions to Emotional Stimuli.

Emotion recognition systems aim to develop tools that help in the identification of our emotions, which are related to learning, decision-making and treatment and diagnosis in mental health contexts. The research in this area explores different topics ranging from the information contained in different physiological signals and their characteristics, to different methods aiming feature selection and classification tasks. This work implements a dedicated experimental protocol, consisting of sessions to collect physiological data, such as the electrocardiogram (ECG), while the participants watched emotional videos to provoke reactions of fear, happiness, and neutrality. Data analysis was restricted to features extracted directly from the ECG, being possible to verify that the intended stimuli effectively provoked variation in the heart rhythm and other ECG features of the participants. In addition, it was observed that each emotional stimulus presents different degrees of reactions that can be clearly distinguished by a clustering procedure.