Evaluating Online Articles for Misinformation using Machine Learning and Systematic Data Collection

This project was completed as an entry into the 2020 Synopsys Science Fair. It aims to tackle a portion of online misinformation by combining several methods. The design provides a sentiment analysis of the headline, and returns an article source’s political bias (from allsides.com). The program also lists related articles to widen a consumer’s perspective, as well as a coverage report for viewers to see who is reporting on certain issues.

This was achieved using a flask backend in conjunction with a simple HTML/CSS frontend. The backend handled database and NLP api queries related to the current article the user is reading.