Fake news classifier

Bot application for classifying news articles as REAL or FAKE using Machine Learning

Table of Contents

• About the Project
  • Developed Using
• Demonstration
  • Prerequisites
  • Demo
• Contributing
• References
• Contact

About The Project

DeceptiBOT uses a machine learning model to classify news articles in a dataset as real or fake via the PassiveAgressiveClassifier. This model is then applied to user-submitted-news using a Telegram messenger bot.

Developed Using

This project was developed using multiple modules, documentation can be found below:

• scikit-learn
• Matplotlib
• pandas
• NumPy
• Anaconda
• Pickle
• PyTelegramBotAPI

Demonstration

Prerequisites

First, install all dependencies listed above / in the jupyter file.

Instructions on how to set up Anaconda and the notebook environment can be found here

Instructions on setting up a Telegram bot can be found here

Demo

Head over to any news source and copy the text:

Use the /Start command to initiate the process:

The bot displays its prediction as follows:

Contributing

Have a burning idea you feel could improve DeceptiBOT?

1. Fork the project
2. Create a new branch (git checkout -b update)
3. Commit your changes (git commit -m 'Description of update')
4. Push to the branch (git push origin update)
5. Open a new pull request

References

• Thanks to Data Flair for the ML tutorial
• Credit to othneildrew for the README template

Contact

Abhinav Ramesh - abhinavramesh03@gmail.com