Text-Based Cyberbullying Prevention using Toxicity Filtering Mobile Chat Application and API

Abstract

The internet becoming a ubiquitous thing for most people has both positive and negative consequences. One negative consequence is that everyone’s profile or contact info on any social media is made available to anyone on the internet. With the ever-growing userbase of social media platforms, the risk of being cyberbullied is very high. Since most communication on any social media platform is done through chats, an attempt has been made to curb the cyberbullying on these social media platforms in textual form. This will be done by providing an API (Application Programming Interface) that can receive an input text and respond with an annotation if the text is predicted to be offensive or not and a framework for supporting the same algorithm and running the artificially intelligent model that can understand natural language on mobile devices locally to offer a complex service to the end-users directly without having to depend on the internet and compromising on privacy. Finally, making this available in form of a mobile application would give a lot of user’s access to an extremely useful and helpful system.