Detection of Fake News Using Binary Classification

Abhyudya Bajpai, Amit Kumar Tripathi

Abstract: The idea behind this project is to detect the accuracy of the fake news using Binary Classification such as Multinomial Na?ve Bayes, Passive Aggressive classifier. Here the two datasets are provided i.e., test dataset and train dataset. Test data is later matched with groups of train dataset and accuracy is found using Binary classification. This helps in determining whether given news is fake or real. It delivers maximum accuracy and helps to identify fabricated news. The data is pruned by removing stop words and common English words by using vectorizer.

Keywords: Fake news, Binary Classification, Multinomial Na?ve Bayes algorithm, Passive Aggressive Classifier algorithm, outliers, TFIDF Vectorizer