“Machine Learning and Content Coding Study Analyzes Conspiracy Discourse on 5G and COVID-19 on Twitter”

https://pubmed.ncbi.nlm.nih.gov/38117843

This study combines machine learning and qualitative content coding to analyze conspiracy discourse related to 5G and COVID-19 on Twitter, finding that conspiracy-related discourse is more likely to express negative emotions and use combatant language, while corrections discourse is more likely to use prosocial language and reference health consequences.