Fake News Detection using Supervised Machine Learning Algorithm with Feature Extraction Method

Abstract

The Internet community has become the most popular platform on which people share
information with others with ease of access. Some people use this opportunity by sharing
false or fake information, intentionally or unintentionally. Faux news is incorrect or deceptive
data distributed as an article with the intention of tarnishing someone's or an organization's
reputation or making money off of false information. False news is spread via smartphone
instant messengers such as Twitter, Facebook, Instagram, and WhatsApp. Sometimes, fake
news is spread through social media without knowing its authenticity, and it will mislead the
people. To overcome this problem, the first source of the news needs to be identified. Second,
verify the authenticity of the news. This analysis focuses on a feature extraction method for
distinguishing between fake and real news. The fake and real datasets from kaggle are used
with the TF-IDF and countvectorizer feature extraction methods with supervised machine
learning algorithms to improve the efficiency.