September 24, 2025JMIR Formative ResearchOpen Access

Surveillance of Twitter Data on COVID-19 Symptoms During the Omicron Variant Period: A Sentiment Analysis

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Sentiment analysis reveals trends in covid-19 symptoms on twitter, suggesting correlations with omicron variant periods.

Negative sentiment on symptoms like fever and sore throat peaks during omicron variant dominance, indicating possible trends.
The analysis of 391,508 tweets shows significant fluctuation patterns based on emerging omicron subvariants.
Utilizing a fine-tuned RoBERTa model for sentiment analysis achieved 99.7% accuracy in categorizing tweet sentiments.
This study highlights the relevance of real-time social media data in monitoring covid-19 infections and symptoms.

Zhang et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark