Celebrities emerge as advocates in tweets about bullying.

Abstract

To understand the ways that social media users connect with celebrities about bullying, 1,280,151 posts that mentioned one of the top 302 celebrity users named within bullying keyword posts on Twitter between January 1, 2012 and December 31, 2012 were analyzed. Social science and computer science methods were combined to identify how individuals defined celebrities according to bullying roles and what features of celebrities and the tweets were related to the bullying roles. The results show that Twitter users defined celebrities most frequently as potential advocates against bullying. The results of a Principal Components Analysis identified an advocate/confidant role dimension and the distribution of celebrities across these dimensions were associated with celebrity type and age. A sentiment analysis revealed that the sentiments of the tweets that defined celebrities as advocates were more likely to be positive (e.g., "THANKSSSSS FOR YOUR SUPPORT"), than negative (e.g., you are MEAN). Together, the results suggest that social media is a space that holds potential for advocating against bullying versus being perceived primarily as a space where cyberbullying occurs. Knowing that social media is a context where celebrities can act as advocates, defenders, and role-models in the fight against bullying can be used to inform prevention, intervention, and advocacy programs.

OMG U #Cyberbully! An Exploration of Public Discourse About Cyberbullying on Twitter

“…These URL sites provided useful information including support for victims and strategies for dealing with cyberbullying. Similarly, Resnik et al’s (2016) analysis of tweets showed that tweets containing references to celebrities in an advocate role, that is, as someone who might take or who has taken a general antibullying stance, were largely positive in sentiment. These suggest that multiple avenues exist to combat cyberbullying messages.…”

“…Computer scientists have focused on cyberbullying primarily through the development of methods to detect or block aggressive texts as well as user interfaces that encourage reflection by the perpetrator before engaging in aggressive behavior (Al-garadi et al, 2016; Raman et al, 2015). To a lesser extent, researchers have analyzed the content of tweets such as sentiment (Choi et al, 2014; Resnik, Bellmore, Xu, & Zhu, 2016), hashtags (Calvin et al, 2015), and the role of the audience or online bystanders (Cocea, 2016). Previous Twitter research suggests that each approach to text analysis includes both strengths and weakness, for example, machine learning faces challenges when coding slang and “informal text” (Del Bosque & Garza, 2014), while human analysis might include personal bias due to individual experience.…”

Deception strategies and threats for online discussions

“…Such technologies make detection of social bots more difficult and provide an advantage to the bot creators in this arms race. Targeted attacks are made possible through the anonymous use of social media, by orchestrating a large army of social bots, trolls (McCosker, 2014;Aro, 2016), sock puppets, and bullies (Bellmore et al, 2015;Resnik et al, 2016). Examples of extremist activities on social media have been increasing at an alarming rate, and many platforms have started taking precautions for early-detection and prevention of such activities.…”

Automated Detection of Doxing on Twitter

“…We focus on Twitter because it is a straightforward source of large volumes of online social networks data and provides a public REST API for collecting the streams of tweets. 3 While some datasets are available for detecting other types of cyberbullying on Twitter [10,41], they have not focused on doxing and sensitive information disclosure specifically. Therefore, they mostly do not contain any sensitive information, and we had to collect a new set of tweets.…”