100%(1)1 out of 1 people found this document helpful
This preview shows page 6 - 10 out of 26 pages.
found that it would be difficult for social media technology to effectively detect cyberbullying based on specific keywords due to other “factors such as severity, level of insult and duration, and power are yet to be considered in the operational definition of cyberbullying” (Ioannou, Blackburn, Stringhini, De Cristofaro, Kourtellis, & Sirivianos, 2018, p. 261).
Literature Review (continued)Ting, Liou, Liberona, Wang and Bermudez (2017) developed the idea to approach the detection of cyberbullying “based on technique of Social Networks Mining with three important features, including Keywords, SNA measurements and Sentiment” (p.2). Silva, Rich, Chon, and Tsosie (2016) created a parental controlled application via Facebook called BullyBlocker which is notifies parents when cyberbullying occurs. The results of the study identified age and gender as important factors to the various reports of cyberbullied victims. A Facebook Analysis of Helping Behavior in Online Bullyingstudy bases their research on social intervention from online bystanders to prevent or eliminate cyberbullying. The findings from the study suggest human intervention is more effective to prevent “indirect aggression with targeted attacks found in conventional direct aggression” than algorithm produced bots in solving early detection of cyberbullying (Freis & Gurung, 2013, p. 16).
Literature Review (continued)According to Milosevic (2016) “some platforms lend themselves to subtle forms of bullying, while other sites witness ‘more blatant’ bullying. The nature of cyberbullying on these platforms has important implications for how anti-bullying mechanisms were designed and how their effectiveness is conceptualized” (p. 5171). This article is great to use with my research because it addresses different social media company policies in regulating various media such as images, comments, video footage and even live broadcasting from their users. Another supporting research found that “nearly half of students had been trolled within the past six months and each student was trolled more than one time per month” (Case & King, 2017, p. 32).
Literature Review (continued)Upadhyay, Chaudhari, Arunesh & Ghale (2017) illustrated their primary objective was to “to detect such cases over the social media and implement some preventive measures so as to avoid such incidents” (p.1). Upadhyay, Chaudhari, Arunesh and Ghale (2017) suggests in order for social media technology to early detect and prevent cyberbullying, an effective algorithm must “extract keywords from text files containing different categories data and user messages” (p. 3). Detecting Cyber Bullies on Twitter using Machine Learning Techniquessuggests using key features such as user activity to develop a cyberbullying algorithm. This illustrates the importance of using classification within the algorithm to detect cyberbullying.