Improving Multiclass Classification of Fake News Using BERT-Based Models and ChatGPT-Augmented Data | Publicación