Fake News Detection (Team 18)

To Run (training):

• Save the NELA-GT 2019 dataset in Google Drive
• Run Experiments-jupyter/Fake-News.ipynb (change paths to match)

To Run (testing samples):

• Install the requirements from streamlit/requirements.txt
• Save the corresponding testing files from training as joblib files
• Run 'streamlit run visualizer.py' (change paths if needed)

Best Performing Model: TF-IDF Vectors with SVM (90.6%)

Original Notebook links:

• https://colab.research.google.com/drive/1syECF2-gCI3KVg0lMnDL655iULSgyZLS?usp=sharing
• https://colab.research.google.com/drive/1yL2pYjWA_cMxRAB-X9SzZzI65zFyF4I4?usp=sharing
• https://colab.research.google.com/drive/1LvgBbKJGiBkNQXiko6dUg14MHK2K58EK?usp=sharing
• https://colab.research.google.com/drive/1h8B58R4xbpcDr-pZ0uApPD-b-9d5cNvw?usp=sharing