A Systematic Survey on Computational agents for Mental Health Aid

Abstract

In recent times the rate of mental health disorders among the youth has spiked due to increased competitiveness which affects the mental well-being of students. An unhealthy mental state not only affects the daily life of an individual but is also the reason for increased self-harm and suicide rates. In developing nations, the ratio of mental care professionals to mental health patients is far too less for all to receive care. A solution to this problem is internet-delivered cognitive therapy (iCBT). The objective of this paper is to shed light on various techniques that can deliver iCBT to a patient in a comfortable manner. Since iCBT can be delivered from home, it tackles the challenge of societal stigma. Different existing approaches and solutions being implemented like various types of chatbots(SERMO, EMMA) and social robots(Ryan Bot)are analyzed and compared in this paper. We also analyze different types of existing datasets(NHS Mental Health Dataset, CounselChat Dataset, ISEAR Dataset) used to train various models(Convolutional Neural Networks, Recursive Neural Networks, Hierarchical Attention Network, Transformers). Word Count Per Session, Sentiment Analysis and Emotion Analysis were some of the evaluation metrics analyzed.