3rd Place Winner of Spring 2021
by Bald Spots: Anh-Thu Nguyen, Ramya Susarla, Annie Zhu
Abstract: The current political polarization in the United States is often attributed to the immense amount of hate speech online. And with the Unite the Right rally of 2017 and the recent Capital insurrection, the task of detecting hateful remarks has never been of more importance. References and the nuances of the English language bring about a need for context; simply flagging slurs and certain phrases is not enough. Through this semester, we have worked to detect hate speech on twitter by contextualizing and deriving linguistic features from tweets. After experimenting with data cleaning, feature engineering, and various models, we had reached an accuracy of 97.48%