Comparative Analysis of Various Machine-Learning and Deep Learning Model for Hate Speech Detection

Toxic or hate comments, texts, phrases are most common on social media, as internet is freely available in today’s world, everyone have an access to it and can throw shades on anyone across the world without bearing any real-world consequences. Sitting in front of a computer screen and spreading toxic comments to people is very easy in the current world, but this situation has a very dark side which leads to cyber bullying followed by depression and increase in suicide rates, even tech giants are facing huge issues regarding online toxicity. Right now there is no foolproof or solid method to tackle these types of comments, but still with the help of AI and machine learning, stopping cyberbullying is somewhat possible. Our main aim is to create a solution which is more accurate and detect toxic and hateful texts, words, phrases, etc. Multiple machine-learning models are used on the same data set choosing the best performing machine-learning algorithm by comparing their precision, accuracy and other parameters and providing a robust solution which can be integrated with web, application or with any other technology.