Cyberbullying Detection on Multiple SMPs Using Modular Neural Network

As reported in recent statistics, the rate of cyberbullying on social media platforms(SMPs) is increasing daily along with affected population in the teenage and young generation worldwide. There is an urgent need to curb cyberbullying online and help curb its adverse effects on young minds and their psychology. Deep Learning in Artificial Intelligence can prove to be a benefactor in detection of such instances. The existing cyberbullying detection techniques attempt to detect either similar type of bullying or target only one social media platform. The proposed model, hence, will attempt to curb both the given bottlenecks. We propose a model that detects varied cyberbullying instances on multiple SMPs, essentially using individual detection models of three social media platforms Wikipedia, Formspring, Twitter synergically, to make a single prediction.