Classifying illicit dark web content through zero-shot prompting: An empirical study with GPT models | Publicación