We wanted to find fake news and learn valuable skills that will be relevant throughout our lives.

What it does

The program takes in an article, and parses several keywords. It then runs a google search for those keywords, and parses the text from the top several results. Then, the articles are split into sentences, and sentiment analyses are performed on each sentence, with each sentence having a negative, neutral, and positive value. These values are then collapsed into 2 dimensions and plotted, allowing us to detect outliers.

How we built it


Challenges we ran into

Too many to name

Accomplishments that we're proud of

Built our first non-trivial program!

What we learned

Data scraping, PCA, Machine learning, sentiment analysis, html to text parsing

What's next for mailTrophyToRoncalli402

Getting it to predict fake news more accurately, run more efficiently, and parse more articles without being kicked out by google.

Built With

Share this project: