Cyberbullying detection on social multimedia using soft computing techniques: a meta-analysis

Cyberbullying is to bully someone in the digital realm. It has become extremely detrimental as the social media and the internet have become more popular and omnipresent. People use the internet services to viciously attack others from behind a screen. The substantial growth in the dimensionality, heterogeneity, subjectivity and multimodality of social media and the pressing need to timely curtail the damage instigated through cyberbullying, has fostered the need to devise automated mechanisms which detect such unfavorable activities. The use of soft computing techniques to handle such pernicious issue has been studied invariably and widely in literature. This study is to understand the viability, scope and significance of this alliance of using soft computing techniques for cyberbullying detection on social multimedia. This work is a systematic literature review to gather, explore, comprehend and analyze the research trends, gaps and prospects of this pairing in a well-organized way. The contribution of this study is noteworthy as it focuses on the use and application of soft computing techniques for cyberbullying detection on social multimedia utilizing a meta-analytic approach in order to integrate, interpret and critically analyze the findings in the original studies for expounding novel approaches to achieve comparable and effectual results pertaining to the defined research domain. Published studies starting April 2003, accessed from six digital portals (ACM, IEEE, Elsevier, Wiley, Springer and Taylor and Francis) have been reviewed to expound the state-of-art within the domain to give insightsand finally identify the directions of future research.