PGPR

PGPR is a unified Bayesian framework that integrates a machine learning model that takes into account review features with peer grading for predicting review conformity.

Installation

1. Create a virtual environment

We are using python3.6 in our implementation, you can create a virtual environment for PGPR using the following command:

sudo apt-get install python3-venv
python3.6 -m venv env-pgpr
source env-pgpr/bin/activate

2. Install requirements

After cloning this repository, navigate inside it:

git clone https://github.com/eXascaleInfolab/pgpr.git
cd pgpr

Install all requirements using the following command:

pip install --upgrade pip
sudo apt-get install gcc python3-dev
pip install -r requirements.txt

If you would like to use PGPR with a GPU machine, you can install CUDA with the following comman line.

conda install pytorch torchvision cudatoolkit=9.2 -c pytorch

Running PGPR

To predict conformity of reviews from ICLR 2018, you can use the script: iclr_18.sh in the scripts folder

cd code
sh ../scripts/iclr_18.sh

To predict conformity of reviews from ICLR 2019, you can use the script: iclr_19.sh in the scripts folder

cd code
sh ../scripts/iclr_19.sh

Citation

Please cite the following paper when using PGPR:

@inproceedings{arous2021www,
  title={Peer Grading the Peer Reviews: A Dual-Role Approach for Lightening the Scholarly Paper Review Process},
  author={Arous, Ines and Yang, Jie and Khayati, Mourad and Cudr{\'e}-Mauroux, Philippe},
  booktitle={Proceedings of the Web Conference (WWW 2021)},
  year={2021},
  address={Ljubljana, Slovenia},
}