Psyche Conversa - A Deep Learning Based Chatbot Framework to Detect Mental Health State

Mental health is one of the most pressing challenges in today's modern world. Traditional thinking, family pressure, unemployment issues, homesickness, and an unhappy relationship are the most common reasons for mental illness, which may also lead to suicidal attempts. People in impoverished countries do not give this issue enough attention. Furthermore, many are hesitant to seek professional counseling from a psychiatrist to retain their privacy. Monitoring social media activities, in addition to common symptoms, is critical for improving the accuracy of detecting early signs of mental illness because it impacts mental health. So, rather than a psychiatrist, a user-friendly deep learning-based mobile application that will chat with them to support and monitor their social media behaviors to determine their mental health status will be an effective way of handling the situation. This study proposes a deep learning-based chatbot framework to help mentally ill individuals identify their mental health conditions and provide appropriate therapy. This framework comprises three components: a keylogger module, a chat module, and a deep learning-based mental illness detection module. In the background, the keylogger collects data from the user's keyboard to track social media activity. The chatbot converses with users and stores their daily chat history in real-time. Both modules pass the data to a deep learning model to determine mental health conditions. This study also compared the accuracy of multiple deep learning classifiers such as Conv-LSTM and BERT for the Reddit Mental Health Dataset.