.

ISSN 2063-5346
For urgent queries please contact : +918130348310

TO IMPROVE ACCURACY TO DETECT FAKE NEWS IN SOCIAL MEDIA USING RANDOM FOREST COMPARED OVER NAIVE BAYES ALGORITHM

Main Article Content

V. Lakshmi Narayana, A. Gayathri
» doi: 10.31838/ecb/2023.12.sa1.403

Abstract

Aim: To predict accuracy by using Machine Learning Algorithms to find fake news published in Social Media to discover the best accuracy in determining which news is fake and which is true. Random Forest(RF) and Naive Bayes (NB) are two approaches for detecting anomalies. Materials and Methods: The dataset for the false news identification was collected from www.kaggle.com. The two groups are Random Forest (N=10) and Naive Bayes (N=10). By Adding G power and to fix 80% is the minimum power of the analysis and maximum accepted error is fixed as 0.5 with threshold value as 0.0805% and Confidence Interval is 95%. Results: A Novel Random Forest (RF) Detection Algorithm has been found to be useful in detecting fake news. The accuracy of the Random Forest(RF) algorithm is (82.60%), whereas the accuracy of the Naive Bayes technique is (72.40%). These two algorithms are used to improve the detection of fake news. Furthermore, the independent significant value p=0.0414 (p<0.05) was met, i.e. alpha is 0.01 with a 95% confidence level. Conclusion: The Novel Random Forest (RF) Detection Algorithm looks to outperform Naive Bayes when it comes to recognising fake news on social media.

Article Details