-
-
A non biased article
-
A slightly more biased article
-
Title screen
Inspiration
Many news articles about coronavirus made us wonder if there is a bias among these articles.
What it does
This website uses 3 factors to give a news article a measurement of bias
How we built it
Used Python and Beautiful Soup 4 to obtain articles from websites Used nltk with python to process extracted articles Used Django, javascript, HTML, and CSS to build web application.
Challenges we ran into
Extracting article from websites was difficult as different websites had different structures and NewsAPI had a paywall Figuring out balanced criteria for measuring bias was also pretty difficult. Invalid URLs also used to crash websites.
Accomplishments that we're proud of
Sleep. Discovered many APIs that we didn’t end up using, but could be useful for future projects Learned how to create and modify CSV using python. Figured out how to use NLTK for analyzing human language data. Got better at using javascript and HTML.
What we learned
Learned a lot about natural language.
What's next for Bias Index
WE can improve evaluation algorithm by setting up different criteria for articles with different lengths and adding more factors such as how words are connected in a paragraph and average sentence length of the article, and weighed value for words with different meanings. Also, do research on language and be more transparent with judging criteria with users. let users rate our scores to improve future scoring. Display more than one output, display scores with multiple graphs. Make more complicated system to more efficiently parse through and retrieve articles from different websites.
Log in or sign up for Devpost to join the conversation.