Fakes news is a hot topic in the AI world. Fakes news can have high poltical , business and social impact. Be able to fight against fakes news is becoming necessary. Indeed as a common reader or a journalist, to distinguishe good news from fake news and misleading news is sometimes difficult, very time consuming and involves too many effort considering the huge amount of fake news on the social network: We wanted to find an easy application to tackle with this issue.

What it does

It's a chrome extension working on social networks like Facebook or twitter and which runs an AI algorithm that can predict if the news are fake, misleading or clickbait and if the news will bring some interesting and true information.

How we built it

The chrome extension is built on javascript and runs a python script using a flask application.

Challenges we ran into

There were 3 main difficulties : - find a useful and large dataset with labeled news, integrating the frontend and the backend and obtaining significant results

Accomplishments that we're proud of

Building a Machine Learning System to categorize the news into multiple categories, integrating everything and architecting a good model to learn from the news

What we learned

The project let the front-end team learn a lot about how to use flask and javascript to set up the necessary proxy to obtain the information. The back-end team worked in Natural Language Processing and Machine Learning models

Built With

Share this project: