Social media has become an important venue for diverse groups to share information, discuss political issues, and organize social movements. Recent scholarship has shown that the social media ecosystem can affect political thinking and expression. Individuals and groups across the political spectrum have engaged in the use of these platforms extensively, even creating their own forums with varying approaches to content moderation in pursuit of freer standards of speech. The Gab social media platform arose in this context. Gab is a social media platform for the so-called alt right, and much of the popular press has opined about the thematic content of discourses on Gab and platforms like it, but little research has examined the content itself. Using a publicly available dataset of all Gab posts from August 2016 until July 2019, the current paper explores a five percent random sample of this dataset to explore thematic content on the platform. We run multiple structural topic models, using standard procedures to arrive at an optimal k number of topics. The final model specifies 85 topics for 403,469 documents. We include as prevalence variables whether the source account has been flagged as a bot and the number of followers for the source account. Results suggest the most nodal topics in the dataset pertain to the authenticity of the Holocaust, the meaning of red pill, and the journalistic merit of mainstream media. We conclude by discussing the implications of our findings for work in ethical content moderation, online community development, political polarization, and avenues for future research.
Download