Design and Implementation of Machine Learning Algorithms for the Detection of Misinformation on Social Media Platforms - A Review and Ensemble Approach

Aug 3, 2026

Upload by Rowan

The rapid adoption of social media and public news platforms has accelerated information dissemination, alongside the spread of inaccurate or irrelevant content that misleads public discourse. Distinguishing misleading written content from truthful information through automated categorization is challenging even for domain experts, who must evaluate multiple factors for veracity. This study examines textual features for fake-news detection, trains multiple machine learning classification models, and evaluates them on real-world datasets. The proposed ensemble learning method delivers stronger performance than individual models.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages8

Reading Time20 min

AccessFree to download

I. Introduction

II. Literature Review

Machine Learning Approaches

Design and Implementation of Machine Learning Algorithms for the Detection of Misinformation on Social Media Platforms Ms. Supriya Ashok Bhosale¹, Dr. Suresh S. Asole² 1Ph.D. Scholar, Department of Computer Science & Engineering, Dr. A. P. J. Abdul Kalam University, Indore, MP, India 2Research Guide, Department of Computer Science & Engineering, Dr. A. P. J. Abdul Kalam University, Indore, MP, India [1](1supriya.bhosale13@gmail.com)[supriya.bhosale13@gmail.com](1supriya.bhosale13@gmail.com) , [2](2suresh_asole@yahoo.com)[suresh_asole@yahoo.com](2suresh_asole@yahoo.com) Abstract The proliferation of the internet and the rapid adoption of public news platforms, such as Facebook (FB), Twitter, and Instagram, have facilitated an unprecedented level of information dissemination in human history. Social media platforms enable users to create and share vast amounts of information, much of which is inaccurate or irrelevant to the discourse. Categorizing written content as misleading or disinformation algorithmically presents significant challenges. Even domain experts must consider multiple factors to determine the veracity of an item. To detect false news, researchers advocate using machine learning classification techniques. This study investigates various textual features that can distinguish between false and true content. We train multiple machine learning algorithms using diverse integrated approaches and evaluate their performance on real-world datasets. Our proposed ensemble learning method outperforms individual models. Keywords: Fake News, Machine Learning, Deep Learning I. INTRODUCTION Today, people frequently obtain their news from social media platforms like Facebook, WhatsApp, Twitter, and Telegram, often accepting this information without verifying its accuracy or origin. The global ease, affordability, and accessibility of sharing information via social media increase the likelihood of spreading false information. Deliberately disseminating false information can yield financial or other benefits, such as damaging reputations or influencing government policies. Consequently, numerous research initiatives aim to accurately detect fake news and mitigate its detrimental effects. This paper provides a comprehensive review of current methods for identifying fake news in response to these concerns. The advent of the Internet and social media has significantly simplified the collection and analysis of vast amounts of information. Many people now spend a substantial portion of their waking hours on social media, where they share and discuss news with friends and other users. Consequently, individuals increasingly rely on web-based platforms rather than traditional news sources for their news, given the former's ease of sharing and interaction. However, the quality and reliability of news on social media platforms are generally lower than those of traditional news sources. The rise of various social media platforms has benefitted the media industry by enabling publications to provide timely and up-to-date news, enhancing overall media engagement. Online news sources sometimes publish fabricated information intending to influence public opinion for financial or political gain. Persistent dissemination of fake news can harm individuals and organizations, destabilizing the delicate balance of trust in the news ecosystem. Public perception is continuously shaped by the spread of misinformation. Spammers often exploit clickbait and fake news to generate online advertising revenue. Fake news poses a significant challenge for businesses, journalists, and democracies globally, with severe repercussions worldwide. For instance, the false news about U.S. President Barack Obama being injured in an explosion caused a $130 billion drop in the stock market. Similarly, a false report claiming iodized salt could counteract radiation effects after the Fukushima nuclear leak led to a sudden salt shortage in Chinese supermark

FAQs

Why is fake news detection on social media considered difficult?

Social media content quality and reliability are often lower than traditional sources, and verifying veracity requires considering multiple factors beyond surface text patterns.

What approach does the study use for detecting misinformation?

It investigates textual features and trains multiple machine learning classification algorithms, then applies an ensemble learning method to improve results.

How does the proposed ensemble method perform compared with individual models?

The study reports that the ensemble learning method outperforms individual models on evaluated real-world datasets.

Aug 3, 2026

Upload by Rowan