Machine Learning Approaches for Fake News Detection on Social Media: A Review

Sumit Sanjay Maske Independent Researcher, Bloomington, Indiana, United States.

Avalokiteshvara Journal of Artificial Intelligence (AJAI)
Volume 1, Issue 1, March 2025, pp. 26-36
Review Article

---

Abstract: The way people consume news via social media has significant effects on individuals, communities, and organizations. This impacts various areas, including reputation, beliefs, crime rates, and both mental and physical well-being. With these extensive implications, it's crucial to delve into the influence of fake news on social media platforms. Researchers have been investigating the challenges and key findings surrounding fake news detection. This paper aims to lay the groundwork for future studies and organizational initiatives that critically assess the ramifications of misinformation within communities. Several Strategies have been suggested to identify and reduce the spread of fake news online. Research indicates that multimodal approaches those that incorporate various types of data tend to be more effective than methods relying on a single type of data for detecting misinformation. Additionally, including contextual information has proven to enhance the accuracy of systems designed to detect fake news. Scholars are focusing on pinpointing credible statements and examining user interactions to boost the detection of false information. A vital area of research is understanding how fake news circulates through social networks, as well as the connections among those spreading it. By looking into different forms of news, researchers seek to overcome the limitations of current models to create a more effective automated system for identifying fake news. This review serves as a basis for developing improved, more efficient automated systems for spotting misinformation.

Keywords: Fake news detection, online social media (OSM), Naïve Bayes (NB), Logistic Regression (LR), Multimodality, Contextual information.

📄 Download Full PDF