RADsum

A web application for extractive summarization based on SummaRuNNer.

Paper: https://adrienguille.github.io/publications/ecir2024.pdf
Model: https://github.com/Baragouine/radsum

Run on local

Clone project

git clone https://github.com/Baragouine/radsum_app.git

Create env (conda)

conda create --name RADSUM_APP python=3.9

Activate env

conda activate RADSUM_APP

Enter into back-end directory

cd radsum_app/back/

Install dependencies

pip install -r requirements.txt
conda install pytorch-geometric -c rusty1s -c conda-forge

Install nltk data

To install nltk data:

• Open a python console.
• Type import nltk; nltk.download().
• Download all data.
• Close the python console.

Install Django REST framework and Django cors header

pip install django
pip install djangorestframework
pip install django-cors-headers

Enter into back-end directory

cd back/ # location: radsum_app/back/back

Download data

Download https://drive.google.com/file/d/1NfVJa21NIGyX9USrnS-hpqnaWitORqP0/view?usp=sharing, extract and copy it to radsum_app/back/back.

The path must be radsum_app/back/back/data/.

Migrate

python manage.py migrate

Run back-end

python manage.py runserver

Run front-end

Open a new terminal.

npm must be installed (version 9.8.0 recommended).

Enter into cd radsum_app/front/radsum/.

Run npm install.

Run npm install react-scripts.

Finally, run npm start.

License

© 2023 Raoufdine SAID & Adrien GUILLE - All rights reserved. This software is distributed under the MIT License.