A Deep Learning Framework Using Data Augmentation for Accuracy Improvement to Analyze Users Posts on Social Media to Find Signs of Mental Illness

Abstract

Through their posts on social media, users frequently express their feelings. A deep learning model was developed for this study to determine a user’s mental state based on the data they posted. We gathered articles for this purpose from Reddit forums dedicated to mental health. Our suggested model may disorder, anxiety, depression and Schizophrenia by examining  and learning posting information published by users. Based on their posts, we think our algorithm can help identify people who could be experiencing mental illness. The consequences of this model, which may be used in combination to other methods to track the mental health of individuals who use internet extensively, are also discussed in this paper.