Identifying Dominant Emotion in Positive and Negative Groups of Navarasa Using Functional Brain Connectivity Patterns

Abstract

Natyashastra is an ancient wisdom of Indian performing arts, written by Bharat Muni, and remarkable for its Rasa theory. Rasa means nectar, the sentiments or emotion felt by the spectator while watching a performance. There are nine Rasas (nine emotions), famously known as Navarasa. Eminent scholars have contemplated Rasa as a superposition of several emotions with the dominance of a particular emotion. Empirical brain research may provide evidence to understand the previous theoretical work. Hence, we carry out this research to understand the most dominant Rasa in positive and negative emotion Rasa groups. By dominance, we mean how well the Rasa of the positive emotion group was distinguished from negative Rasa and vice versa. Our analysis is based on EEG data collected on participants while watching movie clips based on these Rasas with capturing time-varying activity using three functional connectivity metrics. Network properties are extracted from networks and utilized to feed as features for Random Forest classifier. We obtained maximum accuracy (greater than 90%) in five pairs between negative and positive emotions. We find the two most dominant Rasas are Sringaram (Love) and Bibhatsam (Disgust), representing positive and negative emotions, respectively. We observe that weaker connections in delta and gamma bands with the lowest network feature values significantly aid in classifying emotions. The strongest connections of delta and gamma connections involve inter-hemispheric and intra-hemispheric engagement patterns respectively, which suggest global and local information processing while watching emotional clips. Beta waves generate strong connections across regions, which suggest inline findings with previous works on beta for the western classification of emotions.