Tweeted Fact vs Fiction: Identifying Vaccine Misinformation and Analyzing Dissent

Abstract

This paper introduces a comprehensive knowledge extraction and management framework aimed at tackling COVID-19 vaccination misinformation. The framework employs advanced natural language processing techniques, such as sarcasm detection and clustering, to analyze and categorize misinformation from online sources, requiring significantly fewer labeled tweets compared to existing methods. It effectively categorizes vaccine dissenting tweets, enhances discourse analysis, and outperforms baseline models in classification tasks.

Shreya Ghosh

Postdoctoral Researcher (2021-2023)

My research interests include spatio-temporal data analytics and NLP (social media data analysis, low-resource language).