Code and article for Journal of Education Data Mining
$ virtualenv -p /usr/bin/python3.7 venv_convincingness
$ source venv_convincingness/bin/activate
(venv_convincingness) $ pip install pip-tools
(venv_convincingness) $ pip-sync requirements/requirements.txt
(venv_convincingness) $ cd code
(venv_convincingness) $ jupyter notebook BERT_Fine_Tuning.ipynb
The above Bert Fine tuning needs to have access to a GPU. The results will be placed in a directory called tmp.
To run all other models and reproduce graphs in article:
(venv_convincingness) $ jupyter notebook convincingness.ipynb