Instructions

1. Download the Stance Classification Dataset
2. Create the directory data/ in this repository. Extract the four topics of CreateDebate to data/CreateDebate/ and both discussions and annotations in the IAC to data/fourforums/
3. Create the directories img/experiment_1/ img/experiment_2 and replicate the structure of data/CreateDebate/ in both
4. If one does not have the nltk corpus downloaded or one is running a fresh kernel, run prerequisites.py
5. Run run_experiments.py