DeepCheck

Inspiration

The COVID-19 crisis led to a lot of pain, death, and isolation all around the world. To fight against the pandemic the Pan-European Hackathon took place. We thought it would be a great opportunity to make a positive impact. So it all started with our brainstorm for the hackathon. What problems come with such a crisis and where can we help? We collected our ideas and coordinated our roles to work as fast and as efficiently as possible.

What it does

In order to mitigate fake news and to detect websites and newspapers which deliberately publish misleading information, we developed DeepCheck. Simply copy and paste an URL of a newspaper article into the search bar. DeepCheck will then give you an evaluation regarding the trustability and credibility of the analyzed article or paper.

How we built it

The application uses a web parser to scrape the contents of news articles. The scraper is based on python. After we obtain the title and body of the article our data model calculates N confidence. N represents the probability in percent whether the article is possibly fake. Our fake news classifier is based on Natural Language Processing (NLP), the Naive Bayes model has been trained with over 6000 objects from our dataset.