EEG-based emotion recognition in Chinese emotional words

Abstract

Emotion recognition is becoming a more and more hot topic nowadays. Many researches on facial recognition and speech recognition have been done. In this paper, we develop a method to facilitate emotion recognition with electroencephalographic (EEG) signals. The main objective of our study is to reveal the relation between emotional responses and EEG data gathered through giving seven subjects word-imagination stimuli. To induce emotions we choose 20 Chinese emotional words which have relatively high emotional intensities. We use support vector machine (SVM) and linear discriminant analysis (LDA) to recognize emotion, respectively, the average recognition rates of 48.78% and 57.04% could be achieved.