Classification of Real and Fake News Using the Artificial Neural Network Method

Abstract

The rapid spread of fake news through digital platforms poses a significant challenge in the current era, as misinformation can quickly influence public perception and disrupt social stability. This research focuses on the development and implementation of an Artificial Neural Network (ANN) model to classify news articles as either real or fake. The study utilizes news data collected from the past year, which is processed through several stages including data preprocessing, feature extraction, model training, and evaluation. The ANN model is designed to recognize complex patterns in news text, enabling it to distinguish between authentic and fabricated information with high accuracy. Experimental results indicate that the proposed ANN approach achieves an accuracy rate of 98% in classifying news articles. The findings also reveal annual fluctuations in the distribution of fake and real news, with a notable decline in fake news observed in 2025, which may reflect increased public awareness and improved verification practices. The study highlights the ongoing risks posed by digital misinformation and underscores the importance of integrating ANN-based solutions into information systems to enhance digital literacy and support more reliable news dissemination. This contribution is expected to assist stakeholders in mitigating the negative impacts of fake news and fostering a more trustworthy digital information environment.