This study analyzes how Twitter’s suspension mechanism relates to policy violations amidst the Russo-Ukrainian War discourse, using a comprehensive dataset of 107.7 million tweets.
Research Scope
Our analysis examines the relationship between content policy violations and account suspensions during one of the most significant geopolitical events of recent times.
Key Findings
- Exploitation Patterns: Identified scams, spam, and advertisements exploiting the conflict’s trending topics
- Policy Violations: Documented systematic abuse of crisis-related hashtags and trending topics
- ML Insights: Machine learning models successfully explain suspension triggers and patterns
- Content Analysis: Comprehensive examination of malicious content strategies
Methodology
- Large-scale Dataset: 107.7M tweets analyzed for suspension patterns
- Temporal Analysis: Tracked suspension patterns throughout the conflict timeline
- Content Classification: Identified different types of policy-violating content
- Predictive Modeling: Developed models to understand suspension mechanisms
Impact
This research provides crucial insights into social media platform governance during crisis situations and contributes to understanding how malicious actors exploit global events for harmful purposes.