Retweet Network Research Articles

Attraction to politically extreme users on social media.

Political segregation is a pressing issue, particularly on social media platforms. Recent research suggests that one driver of segregation is political acrophily-people's preference for others in their political group who have more extreme (rather than more moderate) political views. However, acrophily has been found in lab experiments, where people choose to interact with others based on little information. Furthermore, these studies have not examined whether acrophily is associated with animosity toward one's political out-group. Using a combination of a survey experiment (N = 388) and an analysis of the retweet network on Twitter (3,898,327 unique ties), we find evidence for users' tendency for acrophily in the context of social media. We observe that this tendency is more pronounced among conservatives on Twitter and that acrophily is associated with higher levels of out-group animosity. These findings provide important in- and out-of-the-lab evidence for understanding acrophily on social media.

Differentiating broadcast from viral: a causal inference approach for information diffusion analysis

Classifying information diffusion patterns is critical to many information analysis areas, e.g., misleading information detection. However, diffusion pattern classification remains challenging when multiple users are involved. To address this challenge, this study aims to classify how information diffuses, distinguishing between broadcast and viral spreading, solely through the analysis of observational data from retweet networks on X (formerly known as Twitter). In broadcasting, most users directly receive information. However, viral spreading allows users the opportunity to receive information from a variety of sources. Therefore, viral spreading increases the likelihood of identifying misleading information. Existing methods classify diffusion types mainly through structural virality, which relies on the average distance between the users. However, when dealing with diffusion networks involving two or more information sources, these approaches can potentially lead to confusion regarding causality. To tackle this problem, we develop a deterministic causal inference method for categorizing information diffusion types. To the best of our knowledge, this is the first study investigating information diffusion types based on causality. This approach can be used to assess source credibility and assist in detecting misleading information. It can also be extended to other social media platforms.Graphical

Mapping the terrain of hate: identifying and analyzing online communities and political parties engaged in hate speech against Muslims and LGBTQ+ communities

AbstractThis study investigates the impact of X on political discourse and hate speech in Finland, focusing on Muslim and LGBTQ+ communities from 2018 to 2023. During this period, these groups have experienced increased hate speech and a concerning surge in hate crimes. Utilizing network analysis methods, we identified online communities and examined the interactions between Finnish MPs and these communities. Our investigation centered on uncovering the emergence of networks propagating hate speech, assessing the involvement of political figures, and exploring the formation dynamics of digital communities. Employing agenda-setting theory and methodologies including text classification, topic modeling, network analysis, and correspondence analysis, the research uncovers varied communication patterns in retweet and mention networks. Retweet networks tend to be more fragmented and smaller, with participation primarily from far-right Finns Party MPs, whereas mention networks exhibit wider political representation, including members from all parties. Findings highlight the Finns Party MPs' significant role in fostering divisive, emotionally charged communications within politically segregated retweet communities, contrasting with their broader engagement in mention networks. The study underscores the necessity for cross-party efforts to combat hate speech, promote inclusive dialogue, and mitigate political polarization.

The impact of politicians’ behaviors on hate speech spread: hate speech adoption threshold on Twitter in Japan

In this paper, we hypothesized that a leading politician’s behavior signaling support for the ideas of hate groups will increase hate speech by shifting social norms. To examine the hypotheses, in particular, to quantify the social norms shift, the study utilized an adoption threshold measure based on Twitter retweet networks. Our empirical study focused on the hate speech spread effect of an announcement by Yuriko Koike, Governor of Tokyo, declining to participate in a memorial ceremony for Korean massacre victims. The results support the hypothesis: after Koike’s announcement, adoption thresholds of hate speech users were significantly lowered, while the number of hate speech users increased. Further, this study compared the effect of Koike’s announcement to the effect of a North Korean missile launch, a national security threat to Japan. The average hate speech adoption threshold was lower after Koike’s announcement than after the missile launch, suggesting that a leading politician’s behavior could have a greater impact on shifting norms of prejudice than even a nationally threatening event.

A longitudinal analysis of usage patterns, topics, and information dissemination related to five names for cultured meat on social media

Cultured-meat products, which have been hailed for their potential to address multiple drawbacks of traditional meat production, have received regulatory approval in countries including Singapore and the United States and are experiencing rapid market growth. The name of any product could influence public perceptions of it, and thereby affect consumption; and how cultured-meat products should be labeled remains the subject of debate. However, conducting large-scale consumer tests aimed at understanding the association between public perceptions of such products and the various proposed labels/names for them would be time-consuming and expensive. Therefore, this longitudinal research project on how five common cultured-meat labels have been used, and the associations between particular labels and public perceptions, instead relied on 424,382 relevant messages posted or retweeted on Twitter/X between July 2010 and December 2022. This novel approach enabled us to identify a dynamic interplay between label choice and public perceptions, and that each label was associated with a unique set of topics. Also, using social-network analysis, we were able to delineate the structures of cultured meat-related retweet networks and identify the key influencers within them. Our analysis revealed the importance of labeling within the challenging process of arriving at a consensus about what cultured meat should be called.

A structural approach to detecting opinion leaders in Twitter by random matrix theory

This paper presents a novel approach leveraging Random Matrix Theory (RMT) to identify influential users and uncover the underlying dynamics within social media discourse networks. Focusing on the retweet network associated with the 2021 Iranian presidential election, our study reveals intriguing findings. RMT analysis unveils that power dynamics within both poles of the network do not conform to a “one-to-many” pattern, highlighting a select group of users wielding significant influence within their clusters and across the entire network. By harnessing Random Matrix Theory (RMT) and complementary methodologies, we gain a profound understanding of the network’s structure and, in turn, unveil the intricate dynamics of the discussion extending beyond mere structural analysis. In sum, our findings underscore the potential of RMT as a tool to gain deeper insights into network dynamics, particularly within popular discussions. This approach holds promise for investigating opinion leaders in diverse political and non-political dialogues.

The Shapes of the Fourth Estate During the Pandemic: Profiling COVID-19 News Consumption in Eight Countries

News media is often referred to as the Fourth Estate, a recognition of its political power. New understandings of how media shape political beliefs and influence collective behaviors are urgently needed in an era when public opinion polls do not necessarily reflect election results and users influence each other in real-time under algorithm-mediated content personalization. In this work, we measure not only the average but also the distribution of audience political leanings for different media across different countries. The methodological components of these new measurements include a high-fidelity COVID-19 tweet dataset; high-precision user geolocation extraction; and user political leaning estimated from the within-country retweet networks involving local politicians. We focus on geolocated users from eight countries, profile user leaning distribution for each country, and analyze bridging users who have interactions across multiple countries. Except for France and Turkey, we observe consistent bi-modal user leaning distributions in the other six countries, and find that cross-country retweeting behaviors do not oscillate across the partisan divide. More importantly, this study contributes a new set of media bias estimates by averaging the leaning scores of users who share the URLs from media domains. Through two validations, we find that the new average audience leaning scores strongly correlate with existing media bias scores. Lastly, we profile the COVID-19 news consumption by examining the audience leaning distribution for top media in each country, and for selected media across all countries. Those analyses help answer questions such as: Does center media Reuters have a more balanced audience base than partisan media CNN and Fox News in the US? Does far-right media Breitbart attract any left-leaning readers in any countries? Does CNN reach a more balanced audience base in the US than in UK and Spain? In sum, our data-driven methods allow us to study media that are not often collected in editor-curated media bias reporting, especially in non-English-speaking countries. We hope that such cross-country research would inform media outlets of their effectiveness and audience bases in different countries, inform non-government and research organizations about the country-specific media audience profiles, and inform individuals to reflect on our day-to-day media diet.

Wind energy: Influencing the dynamics of the public opinion formation through the retweet network

Wind power is one of the primary renewable sources that are getting increasing support as a means of addressing climate changes. The debate on its use takes place on social media as a fast and unfiltered forum to express one’s opinions. In this paper, we analyse the discussion concerning wind power coming up on Twitter. We measure the influence of opinions (and of the twitterers expressing them) by building a retweet network. A strong concentration of retweets is observed through the use of both the retweet distribution and the Hirschman–Herfindahl Index (HHI): the discussion appears to be driven by a handful of twitterers who are heavily retweeted. We provide a classification of such top influencers by their role in society. A single climate activist is, by far, the top influencer. International and governmental agencies come second, and energy consultants come third. Companies play quite a minor role, on a par with generalist individual twitterers, while scientists are the least-retweeted category. Those top influencers are also constantly present from month to month, while the overwhelming majority of twitterers change from month to month. Leading energy-related events appear to have a short-lived influence on the discussion (two weeks after their taking place).

The Evolution of Polarization in Online Conversation: Twitter Users’ Opinions about the COVID-19 Pandemic Become More Politicized over Time

Political polarization on social media has been extensively studied. However, most research has examined polarization about topics that have preexisting associations with ideology, while few studies have tracked the onset of polarization about novel topics or the evolution of polarization over a prolonged period. The occurrence of COVID-19 provides a unique opportunity to study whether social media discourse about a novel event becomes increasingly polarized along ideological lines over time. This paper analyzes trends in Twitter polarization in relation to COVID-19 and other geopolitical events of 2020. The first two studies use topic analysis to examine the evolving difference over time in discussions of COVID-19 and other topics by liberals and conservatives on social media. COVID-19-related polarization is initially absent but gradually increases over time, in contrast to polarization related to other events. A third study examines structural polarization in retweet networks and finds that the frequency of counterpartisan retweets reduces over time. Across all three studies, we find evidence that Twitter discussion of COVID-19 has become more polarized over time.

Retweet-BERT: Political Leaning Detection Using Language Features and Information Diffusion on Social Networks

Estimating the political leanings of social media users is a challenging and ever more pressing problem given the increase in social media consumption. We introduce Retweet-BERT, a simple and scalable model to estimate the political leanings of Twitter users. Retweet-BERT leverages the retweet network structure and the language used in users' profile descriptions. Our assumptions stem from patterns of networks and linguistics homophily among people who share similar ideologies. Retweet-BERT demonstrates competitive performance against other state-of-the-art baselines, achieving 96%-97% macro-F1 on two recent Twitter datasets (a COVID-19 dataset and a 2020 United States presidential elections dataset). We also perform manual validation to validate the performance of Retweet-BERT on users not in the training data. Finally, in a case study of COVID-19, we illustrate the presence of political echo chambers on Twitter and show that it exists primarily among right-leaning users. Our code is open-sourced and our data is publicly available.

Grounding force-directed network layouts with latent space models

Force-directed layout algorithms are ubiquitously used tools for network visualization. However, existing algorithms either lack clear interpretation, or they are based on techniques of dimensionality reduction which simply seek to preserve network-immanent topological features, such as geodesic distance. We propose an alternative layout algorithm. The forces of the algorithm are derived from latent space models, which assume that the probability of nodes forming a tie depends on their distance in an unobserved latent space. As opposed to previous approaches, this grounds the algorithm in a plausible interaction mechanism. The forces infer positions which maximise the likelihood of the given network under the latent space model. We implement these forces for unweighted, multi-tie, and weighted networks. We then showcase the algorithm by applying it to Facebook friendship, and Twitter follower and retweet networks; we also explore the possibility of visualizing data traditionally not seen as network data, such as survey data. Comparison to existing layout algorithms reveals that node groups are placed in similar configurations, while said algorithms show a stronger intra-cluster separation of nodes, as well as a tendency to separate clusters more strongly in multi-tie networks, such as Twitter retweet networks.

Global Misinformation Spillovers in the Vaccination Debate Before and During the COVID-19 Pandemic: Multilingual Twitter Study.

Antivaccination views pervade online social media, fueling distrust in scientific expertise and increasing the number of vaccine-hesitant individuals. Although previous studies focused on specific countries, the COVID-19 pandemic has brought the vaccination discourse worldwide, underpinning the need to tackle low-credible information flows on a global scale to design effective countermeasures. This study aimed to quantify cross-border misinformation flows among users exposed to antivaccination (no-vax) content and the effects of content moderation on vaccine-related misinformation. We collected 316 million vaccine-related Twitter (Twitter, Inc) messages in 18 languages from October 2019 to March 2021. We geolocated users in 28 different countries and reconstructed a retweet network and cosharing network for each country. We identified communities of users exposed to no-vax content by detecting communities in the retweet network via hierarchical clustering and manual annotation. We collected a list of low-credibility domains and quantified the interactions and misinformation flows among no-vax communities of different countries. The findings showed that during the pandemic, no-vax communities became more central in the country-specific debates and their cross-border connections strengthened, revealing a global Twitter antivaccination network. US users are central in this network, whereas Russian users also became net exporters of misinformation during vaccination rollout. Interestingly, we found that Twitter's content moderation efforts, in particular the suspension of users following the January 6 US Capitol attack, had a worldwide impact in reducing the spread of misinformation about vaccines. These findings may help public health institutions and social media platforms mitigate the spread of health-related, low-credibility information by revealing vulnerable web-based communities.

12-year observation of tweets about rubella in Japan: A retrospective infodemiology study.

Although rubella is an infectious disease that can be prevented by vaccination, there have been periodic epidemics in Japan, mainly among adult males. One of the reasons for this is the lack of interest in vaccination among the target adult male population. To clarify the reality of the discussion about rubella and provide basic resource for enlightening activities for rubella prevention, we collected and analyzed Twitter posts about rubella in Japanese between January 2010 and May 2022. We examined time series, number of tweets per account, tweeted contents, and retweet network. We found that the weekly number of rubella reports and the number of Twitter posts fluctuate simultaneously. During the 2018 rubella epidemic, the number of tweets increased due to the start of the rubella routine vaccination program and the use of cartoons to raise awareness. While 80% of the accounts posted three times or fewer during the period, some accounts posted multiple times per day for more than 12 years. Medical terms such as vaccines and antibodies were frequently used in the tweet contexts. In the retweet activity, a variety of actors, including mass media, medical professionals, and even rubella sufferers, contributed to disseminate rubella-related information.

CONTESTING THE COVID-19 CONSENSUS: CONNECTIVE ACTION OF PSEUDOANONYMOUS ACCOUNTS ON FINNISH TWITTER

Recent years have witnessed invigorated debate about the contradictions of identity-concealing public participation and its influence on democratic deliberation and policy-making. Some have documented the exploitation of misrepresented identities for manipulation, including the manufacturing and disrupting of political discourses. Others argue that pseudonymous speech fosters less conformist and more secure public spaces for critical social commentary and political activism. In this study, we extend this dichotomy by approaching pseudoanonymous online action as connective action; as focused and communicative political activism aimed to contest the authorities’ Covid-19 response during the pandemic in Finland. Our empirical analysis uses the historical Twitter data, over 4.2 million tweets, by pseudoanonymous Twitter accounts predominantly tweeting about Covid-19 (n = 229). Using mention and retweet networks combined with profile metrics and qualitative reading of tweets filtered through the network analysis, we explore the networked activity of these accounts, the strategies they have employed to contest authorities, and their connections to political actors and other more organized forms of mobilization around Covid-19. The preliminary analysis highlights the prominence of these accounts in Covid-19 related Twitter discussions, along with their distinct behavioural features. Pseudoanonymous accounts average shorter lifespans and higher deactivation rates. Some accounts have transitioned from named accounts towards identity-concealing action during the observation period. Further, our analysis reveals a symbiotic relationship between the pseudoanonymous accounts and the established Covid-19 movements. We argue that pseudoanonymous participation is an increasingly emblematic form of connective action afforded by social media platforms, particularly fitted to undermine the legitimacy of authorities.

Information Diffusion or Deviation? An Exponential Random Graph Model of Activism Against Discrimination on Airbnb

This study examines the information diffusion of “Airbnb discrimination” and related digital activism on Twitter. The model-based inferential network analysis shows a hub-and-spoke network that is loosely connected, indicating that the diffusion network of Airbnb discrimination includes various issues (e.g., antisemitism). The retweet network mirrors the existing offline relationship of users (e.g., those who support opinion leaders of antisemitism), demonstrating a core–periphery structure during the information diffusion process. This study not only empirically tested the retweet relationships but also provided theoretical explanations on how these relationships form through the notion of curative logic and two step information flow theory. The findings of this research challenge the widely accepted assumption about social media that “the more users share, the more people will see.”

Homophily and polarization on political twitter during the 2017 Norwegian election

Ideological homophily on social media has been receiving increased scholarly interest, as it is associated with the formation of filter bubbles, echo chambers, and increased ideological polarization. And yet, no linkage necessarily exists between ideological homophily, echo chambers, and polarization. Despite political interactions on social media taking place to a large extent between like-minded individuals, cross-cutting interactions are also frequent. Using Twitter data, we investigated the extent to which ideological homophily, echo chambers, and polarization occur together and characterize the network of political Twitter users during the 2017 election in Norway. Despite the presence of some degree of ideological homophily, we did not find evidence of echo chambers in the Norwegian political Twittersphere during the 2017 election. And yet, the retweet network is characterized by a significant degree of polarization across ideological blocs. Our findings support the thesis according to which polarization on social media may have drivers other than the technological deterministic effect of social media affordances enhancing the formation of online echo chambers.

Protesting is Not Everything: Analyzing Twitter Use During Electoral Events in Non-democratic Contexts

Despite the fact that during political protests, Twitter usage has been substantially studied in various contexts, there are still some significant gaps in our understanding of the ways that this microblogging network is employed in regular political happenings, e.g., elections, particularly in authoritarian countries. As a result, it remains unclear if citizens in non-democratic countries use Twitter to protest at the time of regular political events as the time of uprisings or not. This investigation tries to address this gap by providing some empirical evidence from the Iranian Twittersphere during the 2017 presidential election. Having employed networked framing theory, we combined textual and network analytic approaches to investigate a sample of 10,416 tweets of the most influential users in the retweet (RT) network. Findings demonstrate that Iranian users did not significantly challenge the regime and power relations in Iran. They framed the election in a non-critical way dealing with routine political and election frays and debates. They also preferred to attack politicians rather than discuss contentious and deliberate politics. Thus, this research reveals that Twitter is not always a tool for protesting against non-democratic regimes. At the time of electoral events, it could be used as an ordinary communication platform.

Who polarizes Twitter? Ideological polarization, partisan groups and strategic networked campaigning on Twitter during the 2017 and 2021 German Federal elections 'Bundestagswahlen'

Political campaign activities are increasingly digital. A crucial part of digital campaigning is communication efforts on social media platforms. As a forum for political discourse and political communication, parties and candidates on Twitter share public messages and aim to attract media attention and persuade voters. Party or prominent candidate hashtags are a central element of the campaign communication strategy since journalists and citizens search for these hashtags to follow the current debate concerning the hashed party or political candidate. Political elites and partisans use social media strategically, e.g., to link their messages to a broader debate, increase the visibility of messages, criticize other parties, or take over their hashtags (hashjacking). This study investigates the cases of the most recent 2017 and 2021 German federal elections called 'Bundestagswahlen'. The investigation (1) identifies communities of partisans in retweet networks in order to analyze the polarization of the most prominent hashtags of parties, 2) assesses the political behavior by partisan groups that amplify messages by political elites in these party networks, and 3) examines the polarization and strategic behavior of the identified partisan groups in the broader election hashtag debates using #BTW17 and #BTW21 as the prominent hashtags of the 2017 and 2021 elections. While in 2017, the far-right party 'Alternative für Deutschland' (AfD) and its partisans are in an isolated community, in 2021, they are part of the same community as the official party accounts of established conservative and liberal parties. This broader polarization may indicate changes in the political ideology of these actors. While the overall activity of political elites and partisans increased between 2017 and 2021, AfD politicians and partisans are more likely to use other party hashtags, which resulted in the polarization of the observed parts of the German political twitter sphere. While in 2017, the AfD polarized German Twitter, 2021 shows a broader division along the classical left–right divide.

The German coal debate on Twitter: Reactions to a corporate policy process

Phasing out coal is a prerequisite to achieving the Paris climate mitigation targets. In 2018, the German government established a multi-stakeholder commission with the mandate to negotiate a plan for the national coal phase-out, fueling a continued public debate over the future of coal. This study analyzes the German coal debate on Twitter before, during, and after the session of the so-called Coal Commission, over a period of three years. In particular, we investigate whether and how the work of the commission translated into shared perceptions and sentiments in the public debate on Twitter. We find that the sentiment of the German coal debate on Twitter becomes increasingly negative over time. In addition, the sentiment becomes more polarized over time due to an increase in the use of more negative and positive language. The analysis of retweet networks shows no increase in interactions between communities over time. These findings suggest that the Coal Commission did not further consensus in the coal debate on Twitter. While the debate on social media only represents a section of the national debate, it provides insights for policy-makers to evaluate the interaction of multi-stakeholder commissions and public debates.

Twitter sentiment analysis for COVID-19 associated mucormycosis.

Purpose:COVID-19-associated mucormycosis (CAM) was a serious public health problem during the second wave of COVID-19 in India. We planned to analyze public perceptions by sentiment analysis of Twitter data regarding CAM.Methods:In this observational study, the application programming interface (API) provided by the Twitter platform was used for extracting real-time conversations by using keywords related to mucormycosis (colloquially known as “black fungus”), from May 3 to August 29, 2021. Lexicon-based sentiment analysis of the tweets was done using the Vader sentiment analysis tool. To identify the overall sentiment of a user on any given topic, an algorithm to label a user “k” based on their sentiments was used.Results:A total of 4,01,037 tweets were collected between May 3 and August 29, 2021, and the peak frequency of 1,60,000 tweets was observed from May 17 to May 23, 2021. Positive sentiment tweets constituted a larger share as compared to negative sentiment tweets, with weekly variations. A temporal analysis of the demand for utilities showed that the demand was high in the initial period but decreased with time, which was associated with the availability of resources.Conclusion:Sentiment analysis using Twitter data revealed that social media platforms are gaining popularity to express one’s emotions during the ongoing COVID-19 pandemic. In our study, time-based assessment of tweets showed a reduction over time in the frequency of negative sentiment tweets. The polarization in the retweet network of users, based on sentiment polarity, showed that the users were well connected, highlighting the fact that such issues bond our society rather than segregating it.