Sustainable supply chain security through BEART-based fake news detection on supplier practices

In the context of sustainable supply chain management, the dissemination of fake news can significantly undermine supplier credibility and disrupt operations. This paper presents a comprehensive analysis of a deep learning-based model for Fake News Detection in Sustainable Supply Chain Management. Leveraging the BEART algorithm, the model achieved superior accuracy and robustness compared to traditional machine learning approaches. Extensive experiments across multiple epochs demonstrated consistently high accuracy and low loss during both training and testing phases, with an accuracy rate of approximately 99.9%. The classification report and confusion matrix confirmed the model’s near-perfect precision and recall.