FAKE NEWS DETECTION USING AN ENHANCED SUPPORT VECTOR MACHINE WITH SENTIMENT ANALYSIS

Abstract

ABSTRACT Fake news is anything but a new idea, yet it is a usually happening wonder in current occasions. The outcome of phony news can go from being simply irritating to affecting and deceiving social orders or even countries. In previous literature, comparing Support Vector Machine (SVM) and machine learning for text categorization with Sentiment analysis suffers setbacks of low performance and lack in terms of the range of evaluated models and the diversity of the used datasets. The aim of this study is to Enhance Support Vector Machine using Sentiment Analysis for easy detection of rumour on social media platform using individual twitter account. This was achieved by collecting relevant data for performing fake news detection, using SVM and sentiment analysis for easy detection. The results obtained from the study indicate that the technique performed optimally in fake news detection with the accuracy of 98% and a false alarm rate of 0.02. This reveals that the enhancement of SVM with sentiment analysis for fake news detection enhances the performance of the detection model.