- Install Scikit-learn and kaggle python packages
- Run notebooks/collect_data.ipynb
- Make folder
dataandsubmissions - Put the test data gold label files as
submission.csvindata/purpose/anddata/influence. - Run notebooks/3C_models-v1.ipynb and notebooks/3C_models-v2.ipynb to get the final results.
Our solutions came 2nd and 3rd on the Citation Context Classification based on influence and Citation Context Classification based on purpose respectively.