mm5839 Using 3 late days Training just run the command prompt, using selected strings in code to control which features. Special futures: I one-hot encoded user data for gender, party, and political ideology. I also used the VAD lexicon to get the average values for each of V,A,D for the words used by the debaters respectively in a debate. I also looked at punctuation, such as the number of question marks, the number of commas, and the number of exclamation points to get a count as linguistic features to include. In each case, I computed two sets of features: one for the con debater, and one for the pro debater. Please reach out to [email protected] if there are any unforeseen problems. I used regex and NLTK to help in implementing the linguistic features.