Fake News Detection using Machine Learning with Feature Selection

Abstract

Social media has become a popular source for receiving news or information in modern society due to its timeliness and easy accessibility for everyone. However, it causes a series of critical issues. Fake news is one of the issues that urgently need to be resolved. Fake news has the capabilities to compromise democracy and the credibility of information. Compared to other malicious threats, fake news is harder to detect because fake news is created to intentionally deceive audiences. Research has been conducted and suggests that machine learning can be effectively utilized to detect fake news. Thus, we propose a fake news detection system using a k-nearest neighbors (KNN) machine learning model. By utilizing Genetic and Evolutionary Feature Selection (GEFeS) in the fake news detection system, the highest accuracy achieved in this research is 91.3 %. Additionally, we used the GEFeS identified features and an optimal k value to train and test a quantum KNN (QKNN) to explore how quantum machine learning techniques can be utilized in fake news detection problems. The accuracy achieved by the QKNN model is 84.4 %.