Classification of Covid-19 fake news using machine learning algorithms

Abstract

Fake news is a fabrication of the original news intentionally to deceive readers. Internet and social media help such news to spread widely and affect individuals and society negatively. Because of the lack of control over writing the posts on social media. The spread of this type of news has become much more than before. We present one of the most societal severe affairs for misinformation, especially in the presidential elections and fake news related to health like COVID-19. Therefore, there is a need for machine learning algorithms to detect and classify all types of fake news that is difficult to be detected by a human and experts. In this paper, Covid-19 FNs are detected using the Term Frequency-Inverse Document Frequency (TF-IDF) as features extraction and two machine learning algorithms (SVM, Multinomial Naive Bayes) as a classifier. The results show that the accuracy of the proposed algorithms is equal to 94.83% and 91.38%, respectively. We conclude that using machine learning algorithms can help detect such fake news based on good achieved accuracy. © 2022 American Institute of Physics Inc.. All rights reserved.