Mapping the Pandemics Echo: Dynamic Narrative Detection and Spatio-Temporal Sentiment Modeling of COVID-19 Discourse on Twitter

The research analyzed 2.4 million geolocated tweets from January 2020 to June 2022 to understand the evolution of public opinion during the COVID-19 pandemic. Using a specialized COVID-Twitter-BERT model and dynamic topic modeling via BERTopic, researchers identified fifty distinct narratives with high coherence scores. The analysis revealed three main pandemic phases: in the first quarter of 2020, fears about mask shortages dominated; in 2021, vaccine optimism was followed by polarization; and in 2022, pandemic fatigue emerged. Regional comparisons showed significant differences between countries – US discourse was dominated by debates about freedom versus mandates, while European discussions emphasized collective solidarity. The framework achieved 76% F1-score in sentiment classification.