montycoder0701 / anti-bait

Uses Python ML to detect fake news.

Anti-Bait 📰

Uses Python ML to detect biased news.
The data to train this program is provided in the file news.csv.

Libraries used

• Numpy
• Pandas
• Sklearn
• Streamlit

Local development (Windows)

py -m pip install --user virtualenv

py -m venv env

Authorize access

Set-ExecutionPolicy Unrestricted -Scope Process

.\env\Scripts\activate.ps1

Installing libraries

py -m pip install [library-name]

How to run unit tests

Run the command below.

py -m unittest -v fake-news.py

How to use the UI

Run the command below.

py -m streamlit run gui.py

The UI should be on http://localhost:8501/

File information

• fake-news.py
  This file contains the model to detect biased news.
  When prompted, enter the headline in text form in the terminal to get either "UNBIASED" or "BIASED".

• cm.py
  This file contains the Confusion Matrix showing the validity of the model.

• gui.py
  This file contains the code for GUI showing the model.

anti-bait's Issues

Add exception handling for wrong inputs

Add exception handling

Remove news.csv and add link for data instead.

Streamlit change favicon and title

Add template web interface

https://medium.com/uptake-tech/how-to-create-a-simple-frontend-api-and-model-with-python-vue-js-a51841c66f8a

https://hackernoon.com/how-to-instantly-create-a-front-end-for-your-python-program

Error handling for sentence with just special characters or just numbers

Make user input possible, then print result to terminal

Change fake to biased

Add graphic chart of most common words

Change repo title

Error handling for blank cases

Add better unit tests

Add unit tests

Customize gui- remove green source code

Add longer text to title

More validity testing

Detects majority of headlines as fake

news.csv needs to be updated, sample size of fake news should be decreased

Final release

Deploy on Streamlit

Finalize UI

Choose color template

Squash commits

Add file to show pie chart and confusion matrix