Fake News Identification and Classification Using DSSM

Abstract

Internet social networks are increasingly being used to spread information, including false information, for a variety of political and economic purposes. By applying technologies like artificial intelligence (AI) and natural language processing (NLP), scientists can create systems that can automatically identify fake news. The widespread use of social media has significant effects on business, culture, and society—both positive and negative. However, in order to identify false news, models must generalize the data and compare it with real news, which makes it difficult to find. It is suggested that enhanced recurrent neural networks and a deep structured semantic model can be used to identify a technique for identifying and classifying bogus news messages. With 99% accuracy, the suggested method—which does not require topic knowledge—naturally identifies distinctive characteristics linked to bogus news. The suggested system's performance measuring approach is predicated on sensitivity, accuracy, and specificity.