As responsible citizens, it is important for everyone to keep updated and stay politically active. Nowadays, people rely on media, most notably news articles, to learn about current events and government-related issues. One problem with this outlet is unreliability - media bias. The Lens is a simple solution!

What it does

This application uses information from various news sources and analyzes the text via machine learning to output comprehensive results. Depending on the subject of interest, The Lens determines the percentage of unreliability and allows users to compare this statistic among various news sources.

How We built it

The Lens retrieved tweets from Twitter timelines of various news organizations and then filtered based on the user's topic of interest. We then created an algorithm using natural language processing to analyze the tweets for exaggerated bias. We report an average of the percent bias per news source and give an example tweet for each to demonstrate.

Challenges We ran into

We spent a lot of time trying to determine the best metric to detect exaggeration for our given twitter data. Once we formulated the algorithm, the biggest challenge was adapting it to accurately detect bias. For example, in addition to our sentiment analysis, we had to account for "extreme" words that suggested biased wording.

Accomplishments that We're proud of

It works!

What We learned

On the backend, teammates gained more experience working with natural language processing. In regards to frontend design, teammates learned how to use new libraries and techniques for an effective display of information (i.e. graphs, animations).

What's next for The Lens

We hope to integrate news articles themselves into our program so that users can both check reliability and choose what to read from one application. We also plan to improve our algorithm and possibly expand it to detect where on the political spectrum the article's POV lies.

Share this project: