Ensemble Classifier Based Fake News Identification in Online Social Network

In today’s context, one of the main mediums of news consumption is through social media. It has become a common trend now to produce fake news for speedy propagation and popularity. Thus, it creates a sort of illusion and deception for readers. Amidst the umpty number of researches done, we could say that there is no research that could accurately predict fake news over online. In this study, the ensemble classifier is employed to develop a model for fake news identification in online social networks. This proposed method applies the stacked ensemble classification model, the proposed approach that learns the represented text model and classifies the textual data into real news and fake news. Then, the proposed approach takes the collaborative decision from the classified data that is generated from the multiple base learners using the weight-based ensemble method. The accuracy prediction and performance evaluation of time consumption of detection fake news are consecutively 80 Percentage and 11 ms respectively. Thus, it improves the efficiency and classification accuracy over large-scale social media messages through an efficient sentiment analysis model.