Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Polarization dynamics: a study of individuals shifting between political communities on social media

Federico Albanese, Esteban Feuerstein, Pablo Balenzuela

TL;DR

This paper addresses polarization on social media by analyzing two time snapshots around the 2020 US election using retweet networks. It employs a Stochastic Block Model (SBM) for two-community detection, VADER sentiment for content bias, and topological metrics to characterize user roles. The study identifies a subset of users who switch between opposing communities and finds that these individuals exhibit distinct topological signatures (lower Pagerank, degree, and centrality) and correlated shifts in sentiment toward Donald Trump. The findings illuminate how cross-cutting users contribute to polarization dynamics and provide a framework for detecting and studying community-shifting behavior in political discourse, with implications for depolarization strategies and platform design. The analysis hinges on two time windows, $t_1$ and $t_2$, and leverages a large public Twitter dataset to reveal robust, time-stable patterns of community structure and sentiment.

Abstract

Individuals engaging on social media often tend to establish online communities where interactions predominantly occur among like-minded peers. While considerable efforts have been devoted to studying and delineating these communities, there has been limited attention directed towards individuals who diverge from these patterns. In this study, we examine the community structure of re-post networks within the context of a polarized political environment at two different times. We specifically identify individuals who consistently switch between opposing communities and analyze the key features that distinguish them. Our investigation focuses on two crucial aspects of these users: the topological properties of their interactions and the political bias in the content of their posts. Our analysis is based on a dataset comprising 2 million tweets related to US President Donald Trump, coupled with data from over 100 000 individual user accounts spanning the 2020 US presidential election year. Our findings indicate that individuals who switch communities exhibit disparities compared to those who remain within the same communities, both in terms of the topological aspects of their interaction patterns (pagerank, degree, betweenness centrality.) and in the sentiment bias of their content towards Donald Trump.

Polarization dynamics: a study of individuals shifting between political communities on social media

TL;DR

This paper addresses polarization on social media by analyzing two time snapshots around the 2020 US election using retweet networks. It employs a Stochastic Block Model (SBM) for two-community detection, VADER sentiment for content bias, and topological metrics to characterize user roles. The study identifies a subset of users who switch between opposing communities and finds that these individuals exhibit distinct topological signatures (lower Pagerank, degree, and centrality) and correlated shifts in sentiment toward Donald Trump. The findings illuminate how cross-cutting users contribute to polarization dynamics and provide a framework for detecting and studying community-shifting behavior in political discourse, with implications for depolarization strategies and platform design. The analysis hinges on two time windows, and , and leverages a large public Twitter dataset to reveal robust, time-stable patterns of community structure and sentiment.

Abstract

Individuals engaging on social media often tend to establish online communities where interactions predominantly occur among like-minded peers. While considerable efforts have been devoted to studying and delineating these communities, there has been limited attention directed towards individuals who diverge from these patterns. In this study, we examine the community structure of re-post networks within the context of a polarized political environment at two different times. We specifically identify individuals who consistently switch between opposing communities and analyze the key features that distinguish them. Our investigation focuses on two crucial aspects of these users: the topological properties of their interactions and the political bias in the content of their posts. Our analysis is based on a dataset comprising 2 million tweets related to US President Donald Trump, coupled with data from over 100 000 individual user accounts spanning the 2020 US presidential election year. Our findings indicate that individuals who switch communities exhibit disparities compared to those who remain within the same communities, both in terms of the topological aspects of their interaction patterns (pagerank, degree, betweenness centrality.) and in the sentiment bias of their content towards Donald Trump.
Paper Structure (14 sections, 2 figures, 3 tables)

This paper contains 14 sections, 2 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Data
  3. Ethical Considerations and Data Availability
  4. Methods
  5. Community detection
  6. Sentiment Analysis
  7. Users interactions
  8. Results
  9. Communities in the retweet networks
  10. Sentiment of the communities
  11. Users who change their community
  12. Topological analysis
  13. Sentiment score
  14. Conclusions

Figures (2)

  • Figure 1: Retweet network of the first time period from May 9th to May 16th, $t_1$, (a) and second time period from June 10th to June 16th, $t_2$, (b) during 2020 US presidential election. Each node is a Twitter user and each edge (directed and weighted) represents the retweets between two given users. The nodes are colored depending on its community: the "republican community" users are in red and the "democratic community" users are in blue.
  • Figure 2: Sentiment score of tweets belonging to the republican and the democratic community for the first time period from May 9th to May 16th, $t_1$, (a) and second time period from June 10th to June 16th, $t_2$, (b) during 2020 US presidential election. (c) and (d) show the Bootsrap sample of mean values at the same times.