Harsha1997/Identifying-Hate-Speech-in-Social-Media

Our aim in this project is to identify if a comment is toxic or not and flag toxic comments for removal. Toxic remarks are ubiquitous in Facebook pages, Instagram and YouTube comments, tweets and reddit threads. This will be extremely beneficial to social media companies (Facebook, YouTube, Tiktok) since it bypasses the need for someone to manually scrape out toxic content; our algorithm will help them do it in real-time. Researchers and social scientists can use the algorithm to identify patterns of hate speech and individuals who try to inflict them. Our model should work for generic conversations and using our machine learning algorithms we will eventually be able to build a robust system that can predict the level of toxicity on future unseen instances.

Jupyter Notebook