Bangla Fake News Detection: Machine Learning Perspective

Abstract

Detecting fake news is a challenging job and likewise has tremendous genuine political and social effects. Mostly it’s spread over social news sites and social media. Fake news is written intentionally to mislead readers which is very bad for society, a country, and also for the whole of mankind. In our mother tongue, fake news spread a lot. Which makes the situation worst? To reduce fake news basically in the Bangla language and ensuring more online and social security we work on this thesis. Because of Fake news, people become misguided and make mistakes. That is why we discuss away in our research to detect Bangla fake news from online news using the advantages of “Machine Learning and Natural Language Processing (NLP)”. We gather around nine thousand Bengali news to get the result. We utilize basic and diligently chose Bangla news to precisely recognize fake news. The test results show a 92.6% accuracy utilizing the SVM model.