Abstract: In order to study the causal effects between brain regions and channels in the process of emotion generation, based on the partial Granger and new causality, a partial new causality (PNC) method for studying causality between time series is proposed. 8 channels in the brain were selected under different emotions, PNC was used to calculate the causal connection between the channels in the brain, and a causal network was built based on the connection. The information flow and betweenness attributes of the nodes in the causal network were analyzed. The causal connection between the PNC causal network and the Granger causal network was regarded as a feature and sent to the SVM for training classification. The experimental results show that the average recognition accuracy based on the PNC causal network and the Granger causal network are 76.4% and 68.5%, respectively, PNC can be applied to calculate the causal relationship between time series.