Skip to content

Latest commit

 

History

History
13 lines (8 loc) · 864 Bytes

README.md

File metadata and controls

13 lines (8 loc) · 864 Bytes

Fake News Detection for Hindi Language

  1. The 'FakeNews_preprocessing_final.ipynb' processes the raw dataset collected from Boomlive and BBC news.
  2. The 'FakeNews_model_final.ipynb' is the model file which trains on the processed dataset classifies the test set to obtain the accuracy measure.

To run either of the files, upload the python notebooks to Google Colab and run the cells individually. An already processed dataset is saved in github repository and therefore, to run the second colab the first colab(FakeNews_preprocessing_final) need not be run.

Alternatively you can also directly run the following colabs:

Preprocessing: https://colab.research.google.com/drive/1rjIVyGIMbmMwcFSbrTxVKjHgBkFtPGm0?usp=sharing

Model: https://colab.research.google.com/drive/137LTs0mqawQZKzIEUBXodYwdkzMDNdpe?usp=sharing