The online spread of extremist ideas has been a growing problem. Team LEMMA has worked to quantitatively model the spread of extremist ideas over Reddit in order to gain insight into how they may spread. A modest dataset of Reddit comments were manually rated on the level of extremist rhetoric present and used to train a machine learning algorithm to automatically classify large swaths of Reddit data. These ratings were then fit to a predictive agent-based model with the hopes of better understanding past trends and potentially forecasting future spread of extremism.