Classification of Fake News on Facebook a Novel Social Network with K-Means Clustering Approach for Against Principal Component Analysis Method for Better Accuracy

Abstract

To classify the fake news on Facebook using machine learning algorithms with improving accuracy. Materials and Methods: The Fake news classification implemented in the dataset is used to detect the exact real meaning of the content. In this research study the dataset is labelled as title, text, subject and date, these data are applied on the machine learning algorithms such as K-Means (KM) Clustering taken as group-1 and compared with Principal Component Analysis (PCA) algorithm taken as group-2 used for 80 percent of g power value and dataset used for this using 0.05 significant value and 95 percent of confidence interval also the standard deviation and error value are used. Results: The novel social network used for this research study with K-Means clustering and PCA algorithms and predicted best algorithm for better accuracy.