EMET is a neural network used to classify fake social media posts based on information retrieved from traditional news source.
You can find more information in our paper EMET: Embeddings from Multilingual-Encoder Transformer for Fake News Detection. The dataset used can be downloaded on figshare
docker build <emet/fakenewsDetection:latest> .
> sudo nvidia-docker run --tty --interactive --userns=host --volume /home/EMET:/home/usr/work --name EMET emet/fakenewsDetection:latest /bin/bash
python3 run_emet.py -t /dataset/paper_train.pkl -v /dataset/paper_test.pkl
Thank you Jing Yang for your collaboration in solve some mistakes on the network.