Exploring Topics Trending on Twitter Amid the COVID-19 Pandemic

Twitter (X) served as an important venue for public discourse during the COVID-19 pandemic, including conversations occurring alongside major contemporaneous news events. This study descriptively compares prospectively collected COVID-19-centric Twitter data from four news-event days: the 2020 U.S. Presidential Election, the Pfizer-BioNTech vaccine efficacy announcement, the U.S. Capitol takeover, and the opening day of the second Senate impeachment trial of Donald J. Trump. A consistent search strategy using “COVID,” “COVID19,” or “coronavirus” was applied across all four collection periods. Social network analysis, lexicon-based sentiment analysis, word-frequency analysis, term frequency–inverse document frequency (tf-idf), bigram analysis, and word co-occurrence analysis were used to compare network structure and textual patterns. The four topic days differed in user-network organization, sentiment patterns, distinctive vocabulary, and word relationships. Vaccine News displayed comparatively positive sentiment patterns, whereas Capitol Takeover exhibited the strongest negative cumulative AFINN sentiment trend. Event-related vocabulary also emerged within the COVID-19-centric conversations, while Second Impeachment showed less distinct event-specific lexical differentiation. Together, the findings demonstrate the utility of combining complementary computational methods to descriptively characterize short-term social-media discourse across contrasting contemporaneous news contexts.