DeepNews: enhancing fake news detection using generative round network (GRN)

Fake news is a crucial issue in social media that spreads intentionally crafted false content designed to mislead the public over the digital platform. Many communities have proposed fake news detection methods to enhance the identification of false content on internet applications. Instead of using typical deep learning architecture, this article introduces a novel architecture of unsupervised learning called Generative Round Networks (GRN) to deal with fake news problems. GRN is a modification of common GAN architecture with an additional round function at the generator and a discriminator. In this study, we gather a huge dataset of real and fake news to learn the underlying features. We present the evaluation metrics to convince the proposed model performance by calculating evaluation metrics. According to the experimental result, the proposed GRN architecture can be a promising solution for detecting fake news in practical applications.