NewsEye

Reading the news with you, checking the facts for you.

Comment

Inspiration

In the information age, it can be hard to know whether what read is true and unbiased or not. With the media being a huge forum of political discourse, it can often also lean towards certain political inclinations. That's why we wanted to build a browser extension that would help you both fact check and identify biases!

What it does

NewsEye goes through the news you read online with you - and sends controversy-generating key terms to APIs that let you know if the piece of news containing them has false claims, while also identifying possible political slants.

How we built it

We used JavaScript and Python to build the back-end functionality of the extension, that is, to extract the text from news websites whenever users visit them, send API calls, and receive results. Flask was used to bridge the gap between Python and JavaScript. We used HTML to take this information and present our findings to the user in an intuitive, easy-to-follow, and non-intrusive way. As for our API usage, we utilized the Google Fact Check API, and for sentiment analysis, Google Cloud Natural Language API.

Challenges we ran into

Getting started in languages we were not familiar with was pretty challenging! Neither of us had ever built an extension before, so it was definitely a huge learning curve that we were able to make after hours of poring over documentation, testing, failing, and seeking mentor advice. Also, we had some trouble with making API calls - which is a major part of our project's functionality.

Accomplishments that I'm proud of

  • Being able to draw on our technical skills to tackle the social issue of media bias.
  • Using JavaScript for a project for the first time!
  • Facing challenges with persistence.

What I learned

Patience, perseverance, and of course, some gnarly technical skills!

What's next for NewsEye

  • Improving the sentiment analysis capabilities.
  • Broaden the categorization to include different political leanings beyond the bimodal conservative-liberal model.
  • Improve the UI elements on the screen for an easy viewing experience.
  • Being able to provide alternative news articles with opposing and neutral biases in order to help people challenge their assumptions and the biases presented to them.

Built With

Share this project:

Updates