Twitter Hate Speech Detection using Machine Learning

There is an unprecedented rise in hate speech on social media sites like Twitter in recent times. This widespread problem affects users, leads to problems in the real world, and makes it hard to moderate material. This study aims to find good ways to find hate speech so that it doesn't have as much of an effect on online groups and other places. By using cutting-edge algorithms, the project takes a thorough method to finding hate speech. To make a strong and flexible hate speech recognition system, machine learning models and deep learning methods are used. To make sure accuracy and dependability, model performance is carefully checked using a variety of measures. Accuracy, precision, recall, and F1 score are the common measures used here to show how well the model can correctly spot cases of hate speech. a measure of its general selective power that shows how well it works at different levels. As the project comes to a close, the thorough review of hate speech recognition models has given us useful information. Even though progress has been made, problems still exist, especially when it comes to dealing with changing speaking trends on social media. The study shows how important it is to keep researching and developing ways to find hate speech. This will help make content management better in the future, which will make the internet safer. A strong ensemble method called the stacking classifier is also used as part of the hate speech recognition model. It achieves an amazing 100% success. In addition, the Hybrid Approach, which used both LSTM and BiGRU models, showed an impressive 94% accuracy. A front end was built using the Flask framework to make testing easier for people. It has login features to make the Twitter Hate Speech Detection system safer and more trustworthy. This makes sure that users have a smooth and reliable way to rate how well the model finds and stops hate speech on Twitter.