Detecting Fake News With Machine Learning

Abstract

Fake news is increasingly prevalent in our modern digital age. It ranges from misleading writing and disguised opinion pieces to pieces of satire. With the advent of social media and the growth of the internet in the 21st century, the creation, access, and spread of false information are rapidly increasing in volume. Widespread fake news can foster many problems, including misinformed public opinions and dangerous levels of political division. Current methods of tackling fake news revolve around manual reviews, fact-checking organizations, and black-listing unreliable sources. However, many of these tactics are easily exploited or ineffective. This paper applied machine learning on a Kaggle dataset to predict whether an article of news was real or fake. We applied three different classifiers, all yielding promising results.