fake-news-classifier
所属分类：自然语言处理
开发工具：Python
文件大小：0KB
下载次数：0
上传日期：2022-12-08 06:20:45
上 传 者：
sh-1993
说明： 对新闻文章的真实程度进行分类，以发现虚假新闻，
(Classifying the truth level of news articles to detect fake news,)
文件列表:
.idea/ (0, 2021-05-13)
.idea/dictionaries/ (0, 2021-05-13)
.idea/dictionaries/Russel.xml (522, 2021-05-13)
.idea/fake-news-classifier.iml (438, 2021-05-13)
.idea/misc.xml (2124, 2021-05-13)
.idea/modules.xml (292, 2021-05-13)
.idea/vcs.xml (180, 2021-05-13)
.idea/workspace.xml (23352, 2021-05-13)
Report.md (1693, 2021-05-13)
config/ (0, 2021-05-13)
config/__init__.py (0, 2021-05-13)
config/architecture.py (333, 2021-05-13)
config/config.py (709, 2021-05-13)
config/data_loaders.py (136, 2021-05-13)
data/ (0, 2021-05-13)
data/__init__.py (0, 2021-05-13)
data/liar_plus.py (2700, 2021-05-13)
data/liar_plus_default.py (3328, 2021-05-13)
data/preprocessing.py (3810, 2021-05-13)
datasets/ (0, 2021-05-13)
environment.yml (4984, 2021-05-13)
main.py (10037, 2021-05-13)
models/ (0, 2021-05-13)
models/BiLSTM.py (1809, 2021-05-13)
models/__init__.py (27, 2021-05-13)
models/pretrained_bert.py (409, 2021-05-13)
notebooks/ (0, 2021-05-13)
notebooks/data_loading.ipynb (4037, 2021-05-13)
notebooks/pants_fire_viz.png (3339622, 2021-05-13)
notebooks/visualize.ipynb (490009, 2021-05-13)
requirements.txt (2668, 2021-05-13)
utils/ (0, 2021-05-13)
utils/PIL_helpers.py (138, 2021-05-13)
utils/__init__.py (36, 2021-05-13)
utils/train_helpers.py (1363, 2021-05-13)
visualization.py (557, 2021-05-13)
... ...
# Fake news classifier
## Abstract
A fake news detector based on the LIAR-PLUS dataset.
## Getting Started
Create a conda environment with python 3.6.8
```bash
conda create --name pytorch1.1 python=3.6.8
conda activate pytorch1.1
```
Install all dependencies using `conda env create -f environment.yml`.
```bash
cd fake-news-classifier
mkdir datasets/
```
Download the `LIAR_PLUS` dataset into `datasets/LIAR_PLUS` and the pretrained word vectors
`wiki-news-300d-1M.vec` from [fasttext](https://dl.fbaipublicfiles.com/fasttext/vectors-english/wiki-news-300d-1M.vec.zip)
into `datasets/word_embeddings`.
### Training
Train the model (for binary classification)
```bash
python main.py --phase train --binary yes
```
For hex classification use `--binary no`
### Testing
Similar to training, provide the phase and binary state options through the CLI.
```bash
python main.py --phase test --binary yes
```
## References
* [Where is your Evidence: Improving Fact-checking by Justification Modeling](https://aclweb.org/anthology/W18-5513)
* 2015, Conroy, Niall J., Victoria L. Rubin, and Yimin Chen.
["Automatic deception detection: Methods for finding fake news."](https://onlinelibrary.wiley.com/doi/epdf/10.1002/pra2.2015.145052010082)
Proceedings of the Association for Information Science and Technology 52, no. 1 (2015): 1-4.
* [Kaggle - Toxic Comments EDA by jagangupta](https://www.kaggle.com/jagangupta/stop-the-s-toxic-comments-eda)
* [Kaggle - Spooky NLP Modelling by arthurtok](https://www.kaggle.com/arthurtok/spooky-nlp-and-topic-modelling-tutorial)
* [Kaggle - Quora Insincere questions by ziliwang](https://www.kaggle.com/ziliwang/baseline-pytorch-bilstm)
* [Awesome Fake News - GitHub](https://github.com/sumeetkr/AwesomeFakeNews)
近期下载者：
相关文件：
收藏者：