Fake News Detection in Different Applications: State-of-the-art and Future Directions

Abstract

Social media platforms such as Telegram, Twitter, WhatsApp, and Facebook have grown indispensable for information exchange and delivery in recent years. Humans have a tendency to trust any piece of content that circulates on social media without questioning its authenticity or origin. Fascinated people sometimes spread fake news on social media because of its convenience for information sharing. Additionally, it is used for personal advantages, such as amending government policies and defaming eminent personalities. Hence, to avoid the fatal outcome of spreading fake news, different methods are developed for its identification that achieves high accuracy. This paper reviews conventional artificial news detection methods to solve the mentioned issue. The Systematic Literature Review, often known as SLR, is a procedure that assesses review papers connected to detecting fake news. A technique known as unsupervised learning was used for this article review, along with ensemble learning, semi-supervised learning, reinforcement learning, and supervised learning. This methodology's effectiveness is evaluated regarding the precision of the measurements taken. In conclusion, the deep learning model generates practical outputs and outperforms classical machine learning models.