Suicdal or depressive thoughts are detected in journal entries using sentiment analysis. A natural Langage Proccessing(NLP) model analyzes jounral entries entered onto the website, and users showing signs of poor mental health are directed to resources to help them imporve their well-being.
The Colab Notebook where I devloped the model can be accesed here: https://colab.research.google.com/drive/117JIc-V6DViaTNyyO-PcD4zLAVBAXe3q?usp=sharing
(NOTE: Heroku has removed their free tier, so the site showcasing the model no longer works. I will soon migrate this project to a different platform)