Mental Health & NLP
Detecting and measuring depression, loneliness, ADHD, stress, and well-being from language and images on social media and digital platforms. Building equitable models that account for racial and cultural differences.
We develop NLP and machine learning methods to detect, measure, and understand mental health conditions from digital traces – social media posts, images, and smartphone interactions.
Key areas:
- Depression and anxiety detection from social media language and images
- Loneliness measurement using natural language processing
- ADHD characterization through Twitter behavior patterns
- Stress detection and monitoring at individual and population levels
- Racial and cultural equity in mental health AI models
Our work has been published in PNAS, JAMA Network Open, and npj Digital Medicine, and covered by the APA, WIRED, The Atlantic, and NBC News.