The purpose of the project was to explore hate speech detection using a Recursive Neural Network (RNN). The project proceeds in 2 steps: sentiment analysis, then hate speech detection. The plan was such that tweets with negative sentiment from the first model would then be fed into the hate speech detection model. The training data for the sentiment model was manually combined based on datasets that were manually labeled. The training data for the hate speech detection were cleaned and prepared such that contractions were replaced and links were removed.