Determination of Biological Signal for Emotion Identification

In this paper, we suggest the optimal biological parameters from ANS (Autonomous Nervous System) for emotion identification. For this, 25 subjects were supposed to watch some video clips which induced either positive or negative emotion while we were measuring their ECG, GSR, and SKT. The collected data was processed and analyzed to identify emotion of subjects. Here, we used SVM(support vector machine) and Neural Network as a pattern recognition tool. Even though the two results from SVM and N/N method were little bit different in the degree of accuracy, we eventually could select 3 parameters of ANS. So we decided these three parameters as the optimal minimum parameters for emotion identification.