- Automatic identification, clustering, and summarization of diasters
- LLM generated summaries for human-friendly disaster tracking
- Responsive UI for browsing disasters and viewing related posts
Developers: Manya Bondada, Stephanie Li
- Developed with Python and Flask
- Deployed on Render
Developers: Sneha Bista, Lauren Nicolas
- Developed with React.js, Vite, Tailwind CSS, and Radix UI
- Deployed on Vercel
Developer: Katrina Lee
- Developed with various Python machine learning packages
- Trained on CrisisNLP data
- Final deployed model uses a LinearSVC classifier via FastAPI
- Deployed on Render
This project was developed as a group as part of CS 4485 - Computer Science Project at the University of Texas at Dallas.
Firoj Alam, Umair Qazi, Muhammad Imran and Ferda Ofli, HumAID: Human-Annotated Disaster Incidents Data from Twitter, In ICWSM, 2021.