An Analysis of the Effectiveness of the Naive Bayes Algorithm and the Support Vector Machine for Detecting Fake News on Social Media

Fake news on social media and other media platforms is prevalent and a major cause for concern because it has the potential to have detrimental consequences on society and the nation. This study examines machine learning techniques, such as support vector machines and Naive Bayes classifiers, and relevant research to identify bogus news. Naive Bayes Classifier and Support Vector Machine with linear kernel trick are two popular methods that were employed in this experiment as part of the hybrid approach The sample size for each classifier is 41. As a result, the Support Vector Machine's accuracy rate is 96.24% higher than the Naive Bayes Classifier's accuracy rate of 87.80%. There is no statistically significant difference between the two groups, according to statistical analysis and an independent sample T-test with a value of p=0.433 (p>0.05). Humans are unable to recognise all of this false information, hence machine learning classifiers that can automatically identify fake news are required. The results show that in this case, the hybrid technique based on a support vector machine with a linear kernel trick outperforms the Naive Bayes classifier.