aedFaCT (automated expert discovery-based Fact-Checking Tool)

This repository includes the source codes of the aedFaCT Web browser extension presented in the paper titled "aedFaCT: Scientific Fact-Checking Made Easier via Semi-Automatic Discovery of Relevant Expert Opinions" accepted for the MEDIATE 2023 Workshop, co-located with ICWSM 2023.

Prerequisites

Flask==2.2.2
Flask-Cors==3.0.10
newspaper3k==0.2.8
pybliometrics==3.3.0
spacy==3.5.0
stanfordcorenlp==3.9.1.1

Steps for Deploying aedFaCT

• Download the source codes.
• Obtain a Google Custom Search API key, create the custom search engines, and update values in aedFaCT-server/server.py.
• Install Pybliometrics and configure as explained.
• Download "StanfordCoreNLP" and "elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5", from the repository of the SIFRank+ algorithm and save them to aedFaCT-server/util/m_sifrank/.
• Download data.rar and extract to aedFaCT-server/util/
• Deploy the server (aedFaCT-server) on localhost or a remote server and update server IP address and port number in aedFaCT-wbe/manifest.json and aedFaCT-wbe/background.js.

Citation

Please cite the following paper if you benefit from this repository in your research:

@inproceedings{altuncu2023,
  title={{aedFaCT}: Scientific Fact-Checking Made Easier via Semi-Automatic Discovery of Relevant Expert Opinions},
  authors={Altuncu, Enes and 
           Nurse, Jason~R.C. and 
           Bagriacik, Meryem and 
           Kaleba, Sophie and 
           Yuan, Haiyue and 
           Bonheme, Lisa and 
           Li, Shujun},
  year={2023},
  numpages={10},
  doi={10.36190/2023.27},
  booktitle={Workshop Proceedings of the 17th International AAAI Conference on Web and Social Media},
  publisher={Association for the Advancement of Artificial Intelligence},
}