-
-
Similarity table
HackGT 7 Project
We were inspired by problematic articles and ads prefacing the November election.
This project uses beautiful-soup to scrape political headlines from popular UK news sites. The project then uses a universal sentence encoder TensorFlow to compare them. It then uses a clustering algorithm to cluster the news sites based on similarity.
Our biggest challenge was that we had difficult deciding on what an optimal system of distribution of news was. We didn't know the ideal outcome of our algorithm until late in the hackathon. The question of how to optimally distribute news in a democracy is an incredibly difficult question, and our answer still isn't nuanced enough.
Built With
- beautiful-soup
- numpy
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.