Fake News Detection

Minor Project

Instructions

Jupyter Notebook

• Install numpy, pandas, scikit-learn, nltk using conda.
• Download the dataset from here. It's ~70 MB.
• Unzip and extract to folder named 'data-payames' within the root directory of the repo. Rename the file as 'data-payames.csv'
• Download the dataset from here and extract to folder named 'data-umich' within the root directory of the repo.
• The root folder of the repository is supposed to contain a file called model.joblib. This is one of the trained models exported to a file.

Data Collection

• Install newspaper3k, pytz using pip.

Web Interface

• Install Django.
• Go to Fake-News-Detection/files/fakenews directory.
• Run command in shell "python manage.py runserver".
• Copy the address from the terminal that will be displayed and paste in browser.