BiParse is a Bipartisan Parsing extension powered by machine learning. In America, the amount of political division increases as the media feeds us extremely polar perspectives on current events. BiParse hopes to alleviate the problem of confirmation bias by making the public aware of the biases of the articles and blogs they read and suggesting other perspectives on the issues of our time.
These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.
BiParse requires numpy, flask, tensorflow, pickle, requests, requests_html, csv, time, bs4, and urllib. The all the articles dataset from kaggle is also necessary (it should be extracted to the /api/data project folder as individual .csv files). Create this directory if not already present.
Upon installation of BiParse and its prerequisites, run setup.py to build the neural net and training data (setup.py will be very ram intensive for a period of time). Run train.py to train the net using the data generated. After that is completed, run biparse.py in order to run the flask server. You can open the chrome extension contained in the /app directory by enabling developer mode in chrome and opening the unpacked extension.
- Flask - The web server used
- TensorFlow - The machine learning library used
This project is licensed under the MIT License - see the LICENSE file for details