Bangla Fake News Detection Using Machine Learning

Abstract

Nowadays, Social platform has become an exoteric means for audience to devour news. The proliferation of perplexing fact in regular access media outlets such as social media sites, news blogs, and online portals have created it difficult to detect actual news sources, thus increasing the need for computational tools able to provide clear-sightedness into the reliability of social media content. Extending fake news in social media is often higher than traditional news sources. The augmentation of Bangla fake news and its extension on social media has become a main anxiety due to its caliber to make demolishing dominance. Different machine learning intercourse have been endeavor to identify English fake news. But in our survey we bu to detect fake news from online portal through comparing both true and false ne We also implemented some advanced deep learning models (CNN,CRNN,GRU,LSTM) that have shown promising results for detecting fake news. In our research based project we also build two more models Support Vector k- . We do comparison between three models for getting better accuracy. We have applied random forest, SVM and KNN on the same data test set and train set and found that random forest performs far better than the other two models.