Twitter Discourse

Twitter discourse analysis examines the content and structure of online conversations on the platform, aiming to understand communication patterns, sentiment, and the spread of information, including misinformation. Current research focuses on identifying and classifying different types of online conflict (e.g., agonistic vs. hateful), predicting the intensity of hate speech within conversation threads, and analyzing sentiment related to specific events (e.g., pandemics). Researchers employ various machine learning models, including transformer-based architectures like BERT and logistic regression, often incorporating contextual information from entire conversation threads to improve accuracy. These analyses offer valuable insights into social dynamics, public health, and the impact of online information operations, informing content moderation strategies and crisis response efforts.