During this project we:
- Used tools like NetMapper and ORA to create, clean and visualize networks from a dataset containing 315k tweets.
- Performed Leiden grouping to separate the tweets into three major disinformation storylines and identified hashtags commonly associated with each storyline.
- Employed stance detection and measures of centrality to identify bots and key actors continuing to promote disinformation despite it being declared fake by health/government organizations.
This repository includes a Presentation with more details about this project.