Exploring Coronavirus Disease 2019 Vaccine Hesitancy on Twitter Using Sentiment Analysis and Natural Language...

Area of Science:

• Public Health
• Data Science
• Computational Social Science

Background:

• Vaccination is key to controlling the coronavirus disease 2019 (COVID-19) pandemic.
• Vaccine hesitancy, fueled by social media misinformation, poses a significant challenge.
• Monitoring social media sentiment can inform vaccination strategies.

Purpose of the Study:

• To develop and evaluate a real-time big data analytics framework for tracking social media vaccine sentiment.
• To assess the relationship between Twitter sentiment and COVID-19 vaccination rates in the United States.
• To explore the predictive power of social media sentiment on vaccine uptake.

Main Methods:

• Developed a real-time big data analytics framework using natural language processing and sentiment analysis.
• Ingested, processed, and analyzed tweets for sentiment and content themes.
• Utilized lag cross-correlation and Granger causality tests to evaluate sentiment-vaccination rate relationships.

Main Results:

• Observed a widening gap with increased negative sentiment post-vaccine rollout.
• Found that positive sentiment increases preceded vaccination rate increases.
• Demonstrated a consistent correlation between Twitter sentiment scores and vaccination rates (lagging by ~1 week) and showed sentiment may predict uptake.

Conclusions:

• Social media significantly influences public response to vaccination, disseminating both information and misinformation.
• Real-time analysis of Twitter vaccine sentiment can track public opinion and potentially forecast vaccine uptake.
• Findings suggest social media sentiment analysis can guide targeted vaccination campaigns and strategies.