arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Social Sciences · Sociology and Political Science

A Survey on Machine Learning Approach for Fake News Detection on Facebook

Monday Simon, Badamasi Imam Ya’u, Mustapha Abdulrahman Lawal, Abdulsalam Ya’u Gital, Isah Muhammad Lamir, Ismail Zahraddeen Yakubu

· Zenodo (CERN European Organization for Nuclear Research)

See on the map

Abstract

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

More like this