Combating Depression in Students using an Intelligent ChatBot: A Cognitive Behavioral Therapy

Chatbots are special agents that respond with the user in natural language just as a human would reply. Specifically, social chatbots are the ones which establish a strong emotional relationship with the user. The main concept behind this chatbot was to provide mental relief to students who undergo different levels of stress and which can be the onset of an inimical depression. In this paper, we proposed an intelligent social therapeutic chatbot which distributes the text into emotion labels namely, Happy, Joy, Shame, Anger, Disgust, Sadness, Guilt, and Fear. Further, based on the emotion label, it identify the users' mental state such as stressed or depressed using users' chat data. For emotion detection, we deployed three popular deep learning classifiers namely, Convolutional Neural Network (CNN), Recurrent Neural Network (CNN), and Hierarchical Attention Network (HAN). In particular, the proposed methodology of the chatbot is domain specific where through the users' interaction, the chatbot will try to prevent the pessimistic actions and rebuild more constructive thoughts.