Navigating Depressive Discourse: NLP Analysis of Reddit and Twitter

This project explores how users express depression on Twitter and Reddit using Natural Language Processing (NLP) and Machine Learning (ML). By analyzing linguistic variations through Log Odds Ratio and training Support Vector Machine (SVM) models, we aim to improve detect how linguistically people express feelings of depression on social media. Our models, leveraging TF-IDF and Empath features, outperform baseline classifiers, revealing distinct linguistic patterns between platforms. Future work includes enhancing feature extraction and developing mental health support tools based on these findings.