IMPROVED METHOD OF FAKE NEWS DETECTION BASED ON THE USE OF CNN NEURAL NETWORK

Abstract

At present, the Internet is the primary source of information. In recent times, the role of online social media (OSM) has significantly increased, which has both positive and negative consequences. The negative role of OSM is associated with the spread of fake news, which affects people's daily lives, manipulates their thoughts and emotions, changes their beliefs, and can lead to making incorrect decisions. The problem of fake news dissemination on OSM is a global issue, and the development of mechanisms to counter it is a current task.

Various proven approaches to detect fake news exist today. One approach is based on the use of various machine learning (ML) and deep learning (DL) algorithms. Another approach is based on sentiment analysis of news content and emotional analysis of user comments. Research by the authors into other fake news detection approaches, different from the ones mentioned, has led to the conclusion that these approaches are effective and promising in terms of using their potential to develop new models with high performance on different datasets.

This article explores the authors' idea of improving the existing approach to fake news detection by using neural network approaches. The idea is based on enhancing the method of fake news detection by increasing the number of neurons in the convolutional layer and adding a dropout layer to the studied neural network.

The rationale for the idea involved the preliminary accomplishment of the following: formulation of the research problem, functional analysis of machine learning (ML) and deep learning (DL) algorithms, and experiments to assess the effectiveness of the proposed method on different datasets.