Document Type

Honors Project On-Campus Access Only


Advisor: Susan Fox


Women are often the victims of violent threats online, creating an antagonizing and unsafe environment on microblogging platforms such as Twitter. Machine learning techniques have been successfully applied to classify the sentiment of tweets. This project investigates their use in recognizing hateful discourse. After collecting and hand-labeling thousands of tweets, we trained and tested a multilayer perceptron artificial neural network that classifies threats. The model is promising and suggests further investigation is worthwhile.



© Copyright is owned by author of this document