Online Harassment Detection using Machine Learning

Online communication has become more accessible with the recent digitalization techniques. This advancement also comes with its own advantages and disadvantages. The major disadvantage is the occurrence of bullying and online harassment in cyberspace, which is referred to as cyberbullying. A lot of effort has been taken to deal with this problem. The sufferers will only know the extent of their psychological and mental consequences. Several research studies have proposed different techniques for identifying cyberbullying, however the majority of them focus on a specific job, such as text recognition systems by using Machine Learning (ML) techniques or voice recognition systems using Natural Language Processing (NLP) techniques. To overcome this drawback, the proposed research study offers a multitasking approach, which can work with both text and audio data.