Mapping Anti-Vaccine Activism: Semantic Similarity to Access Disinformation Communities in Twitter (X) During the COVID-19 Pandemic in Brazil
Abstract: This article uses the semantic similarity between fake tweets about COVID-19 vaccines in Portuguese to create a graph and identify disinformation communities on Twitter (currently X). All 2,857,908 tweets in Portuguese containing the word $vacina$ (vaccine in Portuguese) were scrapped from October 30, 2020, to May 25, 2021. A BERT-based algorithm was used to identify fake tweets and obtain their cosine similarity. The study identified five main disinformation communities, highlighting central figures and their influence within these groups. Each community had a clear central subject, and four had a well-defined central spreader of disinformation. Seven of the ten most central users were banned from Twitter for violating community guidelines. This work shows that semantic similarity can be a powerful tool to map disinformation communities in social networks.
Paper Type: Long
Research Area: Computational Social Science and Cultural Analytics
Research Area Keywords: human behavior analysis, misinformation detection and analysis, NLP tools for social analysis; quantitative analyses of news and/or social media
Contribution Types: Model analysis & interpretability, Data analysis
Languages Studied: Portuguese
Submission Number: 1137
Loading