Fake News Detection using Hybrid of Deep Neural Network and Stacked LSTM

Abstract

Fake News is one of the major concerns for the world and it could cause many problems like misleading people, the potential to influence people’s opinions, etc. Hence distinguishing fake news is very crucial. Fake news can be detected using artificial intelligence methodologies using a title, body of the news, and other attributes such as credibility of the authors, location, media such as images, audio, and video, etc. In this paper, a hybrid of deep neural networks and stacked LSTM (Long Short-Term Memory) is used to detect fake news. We used Glove 300d as a word embedding layer and implemented the concept of stacked LSTM. The model has been applied to two different datasets. The results are compared with previous models and have found with good performance in comparison to the selected previously existing models.