The digital era has access to a plethora of data in the Fourth Industrial Revolution (4IR, also known as Industry 4.0) period, including Internet of Things (IoT) data, cybersecurity data, mobile data, business data, social media data, health data, etc. Machine learning (ML), a subset of artificial intelligence (AI), is crucial for improving the corresponding smart and automated applications and for conducting intelligent analyses of these data. The prevalence of social networking sites (SNS) and their ease of use have steadily altered how knowledge is produced and disseminated in the modern world. Cheap access to the news does not, however, guarantee that more people will be aware of it. Social networks, in contrast to traditional media outlets, also hasten and widen the spread of material that has been purposefully misrepresented (fake news). Spreading fake news like wildfire has a negative impact on people's attitudes, behaviors, and beliefs, which in turn can gravely undermine democratic processes. The key problem facing researchers today is minimizing the detrimental effects of fake news through early detection and control of extensive diffusion. In this review article, we in-depth examine a wide range of several approaches for the early identification of fake news in the body of existing literature. We specifically look at Machine Learning (ML) models for the detection of fake news on Facebook, including their classification and identification. We conclude by outlining some unsolved research problems.