root-
|-fakenewsnet_dataset
|-gossipcop
|-politifact
|-main.py
|-main_ssl.py
|-main_ttt.py
|-README.md
|-requirements.txt
|-search_ssl_imdb.sh
|-search_ssl.sh
|-search_ttt_imdb.sh
|-search_ttt.sh
torch==2.0.0
python=3.9.6
datasets==2.11.0
jieba3k==0.35.1
nltk==3.8.1
numpy==1.24.2
pandas==2.0.0
huggingface-hub==0.13.4
The file main.py is the code to reproduce the results of the Bert in our report. The file main_ssl.py is the code to reproduce the results of the Bert-SSL in our report. The file main_ttt.py is the code to reproduce the results of the Bert-TTT in our report. Since the dataset is too large to put it in a Github repo, please check this link to download the fake news dataset link. Then unzip it and put it in the path follow the folder structure.
Here is an example to run the experiment for Bert-TTT on the fake news dataset Fake.
python main_ttt.py \
--weight_decay 0.001 --lr 0.01 --weight 0.7 --mask_ratio 0.05 --dataset fake
- Fake is stored in
./fakenewsnet_dataset/. You can run the code to get data in thehttps://github.com/KaiDMML/FakeNewsNet.
@article{shu2020fakenewsnet,
title={Fakenewsnet: A data repository with news content, social context, and spatiotemporal information for studying fake news on social media},
author={Shu, Kai and Mahudeswaran, Deepak and Wang, Suhang and Lee, Dongwon and Liu, Huan},
journal={Big data},
volume={8},
number={3},
pages={171--188},
year={2020},
publisher={Mary Ann Liebert, Inc., publishers 140 Huguenot Street, 3rd Floor New~…}
}
- IMDB is loaded from the HuggingFace datasets library.
You can find the details of it in the website https://huggingface.co/datasets/imdb.