Topic-driven toxicity: Exploring the relationship between online toxicity and news topics

Bubble reachers and uncivil discourse in polarized online public sphere

Early optimism saw possibilities for social media to renew democratic discourse, marked by hopes for individuals from diverse backgrounds to find opportunities to learn from and interact with others different from themselves. This optimism quickly waned as social media seemed to breed ideological homophily marked by "filter bubbles" or "echo chambers." A typical response to the sense of fragmentation has been to encourage exposure to more cross-partisan sources of information. But do outlets that reach across partisan lines in fact generate more civil discourse? And does the civility of discourse hosted by such outlets vary depending on the political context in which they operate? To answer these questions, we identified bubble reachers, users who distribute content that reaches other users with diverse political opinions in recent presidential elections in Brazil, where populism has deep roots in the political culture, and Canada, where the political culture is comparatively moderate. Given that background, this research studies unexplored properties of content shared by bubble reachers, specifically the quality of conversations and comments it generates. We examine how ideologically neutral bubble reachers differ from ideologically partisan accounts in the level of uncivil discourse they provoke, and explore how this varies in the context of the two countries considered. Our results suggest that while ideologically neutral bubble reachers support less uncivil discourse in Canada, the opposite relationship holds in Brazil. Even non-political content by ideologically neutral bubble reachers elicits a considerable amount of uncivil discourse in Brazil. This indicates that bubble reaching and incivility are moderated by the national political context. Our results complicate the simple hypothesis of a universal impact of neutral bubble reachers across contexts.

News Coverage of the COVID-19 Pandemic on Social Media and the Public's Negative Emotions: Computational Study

BACKGROUND

Social media has become an increasingly popular and critical tool for users to digest diverse information and express their perceptions and attitudes. While most studies endeavor to delineate the emotional responses of social media users, there is limited research exploring the factors associated with the emergence of emotions, particularly negative ones, during news consumption.

OBJECTIVE

We aim to first depict the web coverage by news organizations on social media and then explore the crucial elements of news coverage that trigger the public's negative emotions. Our findings can act as a reference for responsible parties and news organizations in times of crisis.

METHODS

We collected 23,705 Facebook posts with 1,019,317 comments from the public pages of representative news organizations in Hong Kong. We used text mining techniques, such as topic models and Bidirectional Encoder Representations from Transformers, to analyze news components and public reactions. Beyond descriptive analysis, we used regression models to shed light on how news coverage on social media is associated with the public's negative emotional responses.

RESULTS

Our results suggest that occurrences of issues regarding pandemic situations, antipandemic measures, and supportive actions are likely to reduce the public's negative emotions, while comments on the posts mentioning the central government and the Government of Hong Kong reveal more negativeness. Negative and neutral media tones can alleviate the rage and interact with the subjects and issues in the news to affect users' negative emotions. Post length is found to have a curvilinear relationship with users' negative emotions.

CONCLUSIONS

This study sheds light on the impacts of various components of news coverage (issues, subjects, media tone, and length) on social media on the public's negative emotions (anger, fear, and sadness). Our comprehensive analysis provides a reference framework for efficient crisis communication for similar pandemics at present or in the future. This research, although first extending the analysis between the components of news coverage and negative user emotions to the scenario of social media, echoes previous studies drawn from traditional media and its derivatives, such as web newspapers. Although the era of COVID-19 pandemic gradually brings down the curtain, the commonality of this research and previous studies also contributes to establishing a clearer territory in the field of health crises.

Offline events and online hate

Online hate speech is a critical and worsening problem, with extremists using social media platforms to radicalize recruits and coordinate offline violent events. While much progress has been made in analyzing online hate speech, no study to date has classified multiple types of hate speech across both mainstream and fringe platforms. We conduct a supervised machine learning analysis of 7 types of online hate speech on 6 interconnected online platforms. We find that offline trigger events, such as protests and elections, are often followed by increases in types of online hate speech that bear seemingly little connection to the underlying event. This occurs on both mainstream and fringe platforms, despite moderation efforts, raising new research questions about the relationship between offline events and online speech, as well as implications for online content moderation.

How the term “white privilege” affects participation, polarization, and content in online communication

The language used in online discussions affects who participates in them and how they respond, which can influence perceptions of public opinion. This study examines how the term white privilege affects these dimensions of online communication. In two lab experiments, US residents were given a chance to respond to a post asking their opinions about renaming college buildings. Using the term white privilege in the question decreased the percentage of whites who supported renaming. In addition, those whites who remained supportive when white privilege was mentioned were less likely to create an online post, while opposing whites and non-whites showed no significant difference. The term also led to more low-quality posts among both whites and non-whites. The relationship between question language and the way participants framed their responses was mediated by their support or opposition for renaming buildings. This suggests that the effects of the term white privilege on the content of people’s responses is primarily affective. Overall, mention of white privilege seems to create internet discussions that are less constructive, more polarized, and less supportive of racially progressive policies. The findings have the potential to support meaningful online conversation and reduce online polarization.

Role-Aware Information Spread in Online Social Networks

Understanding the complex process of information spread in online social networks (OSNs) enables the efficient maximization/minimization of the spread of useful/harmful information. Users assume various roles based on their behaviors while engaging with information in these OSNs. Recent reviews on information spread in OSNs have focused on algorithms and challenges for modeling the local node-to-node cascading paths of viral information. However, they neglected to analyze non-viral information with low reach size that can also spread globally beyond OSN edges (links) via non-neighbors through, for example, pushed information via content recommendation algorithms. Previous reviews have also not fully considered user roles in the spread of information. To address these gaps, we: (i) provide a comprehensive survey of the latest studies on role-aware information spread in OSNs, also addressing the different temporal spreading patterns of viral and non-viral information; (ii) survey modeling approaches that consider structural, non-structural, and hybrid features, and provide a taxonomy of these approaches; (iii) review software platforms for the analysis and visualization of role-aware information spread in OSNs; and (iv) describe how information spread models enable useful applications in OSNs such as detecting influential users. We conclude by highlighting future research directions for studying information spread in OSNs, accounting for dynamic user roles.

A Multi-platform Approach to Monitoring Negative Dominance for COVID-19 Vaccine-Related Information Online

OBJECTIVE

The aim of this study was to test the appearance of negative dominance in COVID-19 vaccine-related information and activity online. We hypothesized that if negative dominance appeared, it would be a reflection of peaks in adverse events related to the vaccine, that negative content would attract more engagement on social media than other vaccine-related posts, and posts referencing adverse events related to COVID-19 vaccination would have a higher average toxicity score.

METHODS

We collected data using Google Trends for search behavior, CrowdTangle for social media data, and Media Cloud for media stories, and compared them against the dates of key adverse events related to COVID-19. We used Communalytic to analyze the toxicity of social media posts by platform and topic.

RESULTS

While our first hypothesis was partially supported, with peaks in search behavior for image and YouTube videos driven by adverse events, we did not find negative dominance in other types of searches or patterns of attention by news media or on social media.

CONCLUSION

We did not find evidence in our data to prove the negative dominance of adverse events related to COVID-19 vaccination on social media. Future studies should corroborate these findings and, if consistent, focus on explaining why this may be the case.

Directions in abusive language training data, a systematic review: Garbage in, garbage out

Data-driven and machine learning based approaches for detecting, categorising and measuring abusive content such as hate speech and harassment have gained traction due to their scalability, robustness and increasingly high performance. Making effective detection systems for abusive content relies on having the right training datasets, reflecting a widely accepted mantra in computer science: Garbage In, Garbage Out. However, creating training datasets which are large, varied, theoretically-informed and that minimize biases is difficult, laborious and requires deep expertise. This paper systematically reviews 63 publicly available training datasets which have been created to train abusive language classifiers. It also reports on creation of a dedicated website for cataloguing abusive language data hatespeechdata.com. We discuss the challenges and opportunities of open science in this field, and argue that although more dataset sharing would bring many benefits it also poses social and ethical risks which need careful consideration. Finally, we provide evidence-based recommendations for practitioners creating new abusive content training datasets.

Directions in abusive language training data, a systematic review: Garbage in, garbage out

Data-driven and machine learning based approaches for detecting, categorising and measuring abusive content such as hate speech and harassment have gained traction due to their scalability, robustness and increasingly high performance. Making effective detection systems for abusive content relies on having the right training datasets, reflecting a widely accepted mantra in computer science: Garbage In, Garbage Out. However, creating training datasets which are large, varied, theoretically-informed and that minimize biases is difficult, laborious and requires deep expertise. This paper systematically reviews 63 publicly available training datasets which have been created to train abusive language classifiers. It also reports on creation of a dedicated website for cataloguing abusive language data hatespeechdata.com. We discuss the challenges and opportunities of open science in this field, and argue that although more dataset sharing would bring many benefits it also poses social and ethical risks which need careful consideration. Finally, we provide evidence-based recommendations for practitioners creating new abusive content training datasets.