AI Enhanced Depression Detection Using Chatbot and Facial Data Analysis

Depression affects countless millions and requires early detection and intervention. The approach employed in this study builds an AI driven system for depression detection by integrating, chatbot technology, live video facial analysis and machine learning algorithms that have been trained on a dataset of depression related Twitter tweets. Depression levels are analysed in real time by facial expressions of users and users are engaged through conversations with a chatbot on a platform for expression of emotion. This study will be aimed at gathering and treating analysed referenced dataset from the Twitter and processing it employing deep learning and machine learning techniques which includes supervised techniques, such as, support vector machines (SVM), and neural networks. The performance is evaluated via cross validation in order to guarantee highest accuracy. Besides, the accuracy, precision, recall, and F1score for SVM model are 88.5%, 85.2%, 84.6%, and 84.9%, respectively. At Naive Bayes model, the accuracy of 95.6%, precision of 95.7%, recall of 95.6% and F1 score of 95.6% were attained. Logistic Regression model was best, and it achieved 97.7% accuracy, 97.7% precision, 97.7% recall and 97.7% F1 score.