A Review of AI Techniques for Emotion Recognition in Communication with Applications in Child Safety and Education (Published)
Emotions have a major role to play in the development of human communication, especially in the digital world, where children increasingly communicate through texts and voices. The development of artificial intelligence (AI) to automatically recognize human emotions has tremendous potential to contribute to the safety, education, and mental well-being of children. This paper aims to provide a comprehensive review of various artificial intelligence-based emotion recognition techniques for both textual and vocal communications. The paper discusses various techniques such as Natural Language Processing (NLP) and Speech Emotion Recognition (SER), as well as more advanced techniques such as BERT, RoBERTa, convolutional neural networks (CNN), and self-supervised approaches such as wav2vec. The paper discusses the advantages of developing systems that can integrate both textual and vocal communications to develop a more comprehensive system of emotion recognition. Apart from the development of such systems, the paper also discusses the potential applications of such systems in the development of parental monitoring systems, educational support systems, and mental well-being assessment systems for children. Finally, it outlines future directions for the development of such systems with a focus on benefiting children’s well-being.
Keywords: Artificial Intelligence, BERT, emotion recognition, natural language processing (NLP), parental control, speech emotion recognition (SER)