Inspiration
As social media is becoming more prevalent in people’s lives, people are gaining more information by reading articles that shows up in their social media feeds. However, as social media is programmed to recommend similar posts to what the users have already viewed and share posts which their acquaintances have shown interest, the users are prone to become biased and not well informed in the entire political spectrum. This may be detrimental to society since biased opinions may prevent individuals from making optimal choices for society.
What it does
Our program is designed to rate each Twitter user as a well-informed citizen in the community as well as also give a quantitative measure of their bias towards some political party. When a Twitter name is entered to the program, it returns two ratings: an “informed” rating, and a “bias” rating. The “informed” rating shows how much each user was exposed to any political idea, and how much they are familiar with the current social issues. The “bias” rating shows the tendency of the user in following certain political figures and also shows how informed they are in a certain party’s platform, with a highly positive rating meaning a Republican bias and a highly negative rating meaning a Democratic bias.
How we built it
We created an algorithm to test for how informed and biased each Twitter user is, using the people that he or she follows and comparing them to a list of 2016 presidential candidates' Twitter accounts. We then used Java to create a GUI to display the results.
Challenges we ran into
Our algorithm was difficult to implement because of the large amount of data it requires in order to test. Not only must it process all of the people that the user is following, but also check them against the list of political officials.
Accomplishments that we're proud of
We are proud of learning to use the Twitter API in such a short amount of time.
What we learned
We learned to use the Twitter API in conjunction with Java in order to use and analyze Twitter data. We also learned how to manage our time in order to create a finished product within 24 hours.
What's next for Political Bias Tester
We would like to continue testing Political Bias Tester and increasing its efficiency, as it currently takes an extremely long time to run. We would also like to expand our list of political officials' Twitter accounts, as our list is not yet comprehensive. Finally, we would like to analyze keywords to determine political bias, rather than using only the users that are followed.
Log in or sign up for Devpost to join the conversation.