A systematic review on automated human emotion recognition using electroencephalogram signals and artificial intelligence

Brain-Computer Interaction (BCI) system intelligence has become more dependent on electroencephalogram (EEG)-based emotion recognition because of the numerous applications of emotion classification, such as recommender systems, cognitive load detection, etc. Emotion classification has drawn the recent buzz in Artificial Intelligence (AI)-powered research. In this article, we presented a systematic review of automated emotion recognition from EEG signals using AI. The review process is carried out based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses(PRISMA). After that EEG databases, and EEG preprocessing methods are included in this study. Also included feature extraction and feature selection methods. In addition, the included studies were divided into two types: i)deep learning(DL)-based emotion identification systems and ii) machine learning(ML)-based emotion classification models. The examined systems are analyzed based on their features, classification methodologies, classifiers, types of classified emotions, accuracy, and the datasets they employed. There is also an interesting comparison, a look at feature research trends, and ideas for new areas to study.