A Comprehensive Analysis of Machine Learning Approaches for Fake News Detection and Its Effects

Abstract

This study presents an overview of various studies examining the phenomenon of fake news. It specifically focuses on the impact of fake news on people and the advancements made in machine-learning methods for its detection. The study explores two main strategies for identifying fake news: network techniques and sentiment analysis. It also delves into the use of one-class classification models and the analysis of manually collected datasets from social media platforms. The summary highlights the creation of a labeled benchmark dataset for detecting deception and a study that combines news and social content approaches using machine learning. Additionally, it mentions research conducted on bot-generated posts on Twitter. The primary objective of this new study is to analyze the effects of fake news on individuals and develop machine-learning techniques to detect it, with a specific emphasis on mitigating conflicts, crime, and terrorism resulting from fake news. The methodology involves a systematic literature review to gather qualitative data. The summary concludes by providing a concise summary of the main findings related to the impact of fake news on people and the utilization of machine learning methods for its detection.