ABSTRACT
Recent research highlights the potential of inferring an individual’s mental well-being from their text, particularly on social media plat- forms. While existing work often focuses on binary classification for mental health conditions, our novel project aims to incorporate teenage language nuances—such as puns and dark humor—into the dataset. By leveraging advanced natural language processing (NLP) techniques, we seek to develop a nuanced approach for detecting mental health distress in textual data. This research has significant implications for improving suicide prevention efforts.
The data is available here - https://www.kaggle.com/datasets/nikhileswarkomati/suicide-watch/versions/13