Status: Current Project

Abstract: This project uses computational sentiment analysis and social network analysis to examine how community sentiment and polarization escalate in the lead-up to collective violence. I analyze online discourse, media coverage, and community interactions to identify early warning signals.

Methods: Computational sentiment analysis, social network analysis, time series analysis, machine learning

Data Sources:

• Social media discourse
• Local news coverage
• Community interaction patterns
• Historical violence event data

Key Questions:

• How does sentiment polarization precede violence?
• What role do social networks play in escalation?
• Can computational methods identify early warning signals?

Implications: Develops early warning capabilities for violence prevention while advancing understanding of how digital communication shapes offline collective action.