FakeNews Detection on Twitter
Detect Fake News and measure the credibility level of Covid-19 related news on social media (Twitter) using Natural Language Processing (NLP).
This project includes :
• A comparaison between LSTM & BERT models to classify COVID-19 related news on Twitter (fake or real).
• A BERT model trained on a data-set of +11K tweets, with a test accuracy of 92.73%.
• A web browser plugin that displays the credibility level of tweets.