Inspiration
The inspiration for this app came from the challenge given to us by Cognizant. We feel that Machine Learning and Natural Language Processing can change the way we consume information and make the world a better place by breaking down complicated models into things people can understand and determine to be true or false on their own.
What it does
NewsAPI is a project that loads into the user's browser as a plugin or extension and then when the user interacts with a news website, the news is read in by the Chrome extension and our program then uses Google Cloud's Natural Language Processing to assign a rating to the text based on Sentiment analysis. This then is fed back in through the extension and a new webpage is opened, with ratings for how subjective the article might be, and a highlighted area showing all possible subjective areas of the text.
How I built it
We built this application in three main parts. First, the input, when the user would be viewing certain webpages. Here, we used an API to filter out all the unnecessary parts of the webpage and feed the program the text. Next, we parsed the text into Google Cloud's Natural Language Processing methods which gave us scores based on sentiment. Then, using that score, we calculated subjectivity and opened a new webpage with the news article, but this time with the information given about possible bias and highlighted possible areas of the article that were too biased.
Challenges I ran into
While building this app, we ran into various challenges that required us to look back and rework our approach. We had to switch APIs on the input several times, ran into issues working with the cloud database, and had ongoing issues with the compatibility of the program with the browser packages.
Log in or sign up for Devpost to join the conversation.