Tracing Disinformation with NLP / IN PROGRESS

Overview

This project analyzes the spread of disinformation on social media using natural language processing techniques.

Methods

• Data Collection: Scraped and labeled tweets related to misinformation.
• Modeling: Used BERT-based sentence embeddings to detect narrative structures.
• Impact Analysis: Applied sentiment analysis and spatial statistics to measure influence.

Results

• Identified key misinformation narratives.
• Mapped spread across regions and time.