Fake News Detection Using XLNet Fine-Tuning Model

Abstract

In recent years, the traditional way of getting news from a Television, news paper, or national newscast is gone. Today, online social media provides the fastest news content for people. This, however, brings about the problem of fake news. In fact, fake news detection is one of the challenging tasks in natural language processing to differentiate between real (or true) and fake (or false) news content. In this paper, we propose an XLNet fine-tuning model to predict fake news in a multi-class and binary class problem. Our results show that the proposed XLNet model comparatively achieves a better result than the existing state-of-the-art models.