Detection of fake news using deep neural networks

Abstract

With the increased use of the Internet and social networking sites, information can now disseminate at a rapid rate. It is an age of information where any information can be accessed with a single click. This has increased the risk of the spread of misleading false information. These fake news have negatively impacted people and society. So, a strong mechanism is needed to detect false news and stop its propagation. The content of the news, the source of the news, and the response to the news are the main features that can help to detect the credibility and authenticity of the news. This paper aims to implement deep neural networks to accurately detect fake news. It aims to evaluate the various deep learning models to determine the model that can accurately and efficiently distinguish fake news. The experiment uses the Source Based Fake News (SBFN) dataset, a publicly available dataset for Fake News Detection. Various deep learning models such as Long Short Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers (BERT) along with a hybrid model have been trained and evaluated on this SBFN dataset.