Read Aloud the Text Content

This audio was created by Woord's Text to Speech service by content creators from all around the world.

Text Content or SSML code:

The popularity of online social networks has created massive social communication among their usersand this leads to a huge amount of user-generated communication data. In recent years, Cyberbullyinghas grown into a major problem with the growth of online communication and social media. Cyber-bullying has been recognized recently as a serious national health issue among online social networkusers and developing an efficient detection model holds tremendous practical significance. In this paper,we have proposed set of unique features derived from Twitter; network, activity, user, and tweet content,based on these feature, we developed a supervised machine learning solution for detecting cyberbullyingin the Twitter. An evaluation demonstrates that our developed detection model based on our proposedfeatures, achieved results with an area under the receiver-operating characteristic curve of 0.943 and anf-measure of 0.936. These results indicate that the proposed model based on these features provides afeasible solution to detecting Cyberbullying in online communication environments. Finally, we compareresult obtained using our proposed features with the result obtained from two baseline features. Thecomparison outcomes show the significance of the proposed features.©2016 Elsevier Ltd. All rights reserved.1. IntroductionOnline social networking sites have become immensely popularin the last few years. Millions of users have used these websites asnovel communication tools and as real-time, dynamic data sourceswhere they can create their own profiles and communicate withother users regardless of geographical location and physical limi-tations. In this regard, these websites have become vital, ubiquitouscommunication platforms. The communication data from onlinesocial networks can provide us with novel insights into the con-struction of social networks and societies, which is previouslythought to be impossible in terms of scale and extent. Moreover,these digital tools can transcend the boundaries of the physicalworld in studying human relationships and behaviors (Lauw,Shafer, Agrawal,&Ntoulas, 2010).Cyber criminals have utilized social media as a new platform incommitting different types of cybercrimes, such as phishing(Aggarwal, Rajadesingan,&Kumaraguru, 2012), spamming (Yardi,Romero,&Schoenebeck, 2009), spread of malware (Yang,Harkreader, Zhang, Shin,&Gu, 2012), and cyberbullying (Weir,Toolan,&Smeed, 2011). In particular, cyberbullying has emergedas a major problem along with the recent development of onlinecommunication and social media (O’Keeffe&Clarke-Pearson,2011). Cyberbullying can be defined as the use of information andcommunication technology by an individual or a group of users toharass other users (Salmivalli, 2010). Cyberbullying has also beenextensively recognized as a serious national health problem (Xu,Jun, Zhu,&Bellmore, 2012), in which victims demonstrate asignificantly high risk of suicidal ideation (Sampasa-Kanyinga,Roumeliotis,&Xu, 2014). Cyberbullying is a substantially persis-tent version of traditional forms of bullying with negative effects onthe victim. A cyberbully can harass his/her victims before an entireonline community. Online social media, such as social networkingsites (e.g., Facebook and Twitter) have become integral componentsof a user’s life. Therefore, these websites have become the mostcommon platforms for cyberbullying victimization (Whittaker&Kowalski, 2015), and their popularity and proliferation haveincreased the incidents of cyberbullying (Mark&Ratliffe, 2011).Such increase is commonly attributed to the fact that traditionalbullying is more difficult to practice than cyberbullying, in whichthe perpetrators bully their victims without direct confrontation byusing a laptop or a cellphone connected to the Internet (Kowalski,*Corresponding author.E-mail address:[email protected](M.A. Al-garadi).Contents lists available atScienceDirectComputers in Human Behaviorjournal homepage:www.elsevier.com/locate/comphumbehhttp://dx.doi.org/10.1016/j.chb.2016.05.0510747-5632/©2016 Elsevier Ltd. All rights reserved.Computers in Human Behavior 63 (2016) 433e443