Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

X-posing Free Speech: Examining the Impact of Moderation Relaxation on Online Social Networks

Arvindh Arun, Saurav Chhatani, Jisun An, Ponnurangam Kumaraguru

TL;DR

This study investigates how moderation relaxation on Twitter after Elon Musk's takeover affects hate speech and network dynamics, using a large-scale, time-windowed dataset and a hybrid content-network framework. It introduces Moving PageRank (MPR) to track over-time influence and identifies 57 key actors driving diffusion, revealing that $PR_t(x)$ dynamics correlate with network bridges and cross-community interactions. The analysis shows substantial increases in hate content targeting LGBTQ+ individuals and liberals, shifts in linguistic and semantic patterns toward more explicit hate, and the emergence of dense hate communities connected by bridge nodes. The results underscore the need for balanced platform governance that preserves free expression while implementing countermeasures such as counter-speech and community-driven moderation tools to prevent rapid hate proliferation.

Abstract

We investigate the impact of free speech and the relaxation of moderation on online social media platforms using Elon Musk's takeover of Twitter as a case study. By curating a dataset of over 10 million tweets, our study employs a novel framework combining content and network analysis. Our findings reveal a significant increase in the distribution of certain forms of hate content, particularly targeting the LGBTQ+ community and liberals. Network analysis reveals the formation of cohesive hate communities facilitated by influential bridge users, with substantial growth in interactions hinting at increased hate production and diffusion. By tracking the temporal evolution of PageRank, we identify key influencers, primarily self-identified far-right supporters disseminating hate against liberals and woke culture. Ironically, embracing free speech principles appears to have enabled hate speech against the very concept of freedom of expression and free speech itself. Our findings underscore the delicate balance platforms must strike between open expression and robust moderation to curb the proliferation of hate online.

X-posing Free Speech: Examining the Impact of Moderation Relaxation on Online Social Networks

TL;DR

This study investigates how moderation relaxation on Twitter after Elon Musk's takeover affects hate speech and network dynamics, using a large-scale, time-windowed dataset and a hybrid content-network framework. It introduces Moving PageRank (MPR) to track over-time influence and identifies 57 key actors driving diffusion, revealing that dynamics correlate with network bridges and cross-community interactions. The analysis shows substantial increases in hate content targeting LGBTQ+ individuals and liberals, shifts in linguistic and semantic patterns toward more explicit hate, and the emergence of dense hate communities connected by bridge nodes. The results underscore the need for balanced platform governance that preserves free expression while implementing countermeasures such as counter-speech and community-driven moderation tools to prevent rapid hate proliferation.

Abstract

We investigate the impact of free speech and the relaxation of moderation on online social media platforms using Elon Musk's takeover of Twitter as a case study. By curating a dataset of over 10 million tweets, our study employs a novel framework combining content and network analysis. Our findings reveal a significant increase in the distribution of certain forms of hate content, particularly targeting the LGBTQ+ community and liberals. Network analysis reveals the formation of cohesive hate communities facilitated by influential bridge users, with substantial growth in interactions hinting at increased hate production and diffusion. By tracking the temporal evolution of PageRank, we identify key influencers, primarily self-identified far-right supporters disseminating hate against liberals and woke culture. Ironically, embracing free speech principles appears to have enabled hate speech against the very concept of freedom of expression and free speech itself. Our findings underscore the delicate balance platforms must strike between open expression and robust moderation to curb the proliferation of hate online.
Paper Structure (17 sections, 1 equation, 4 figures, 4 tables)

This paper contains 17 sections, 1 equation, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Dataset
  4. Hateful tweet extraction (D1)
  5. Hateful user timeline collection (D2)
  6. Experiments
  7. How did the Hate Speech landscape change?
  8. Types of Hate Speech
  9. Representative Words
  10. Shifts in Word Semantics
  11. How does the moderation relaxation affect the hate in existing communities?
  12. Can we (early) detect the users who drove the change in this landscape?
  13. Moving PageRank (MPR)
  14. Early Detection of Influential Users
  15. Discussion
  16. ...and 2 more sections

Figures (4)

  • Figure 1: ForestFire subsampled $(|V|=1000)$ visualization of hate interaction network two weeks before and two weeks after the takeover
  • Figure 2: Rate of growth of the average degree centrality of nodes increases by 144.44% post-takeover
  • Figure 3: Rate of growth of the number of connected components decreases by 17.3% post-takeover
  • Figure 4: Spearman correlation $(\rho)$ between MPR rank and user profile metrics for the top 2000 users