Antonio Crudele
Agentic-based Deep Web Crawler for the Detection of Hidden Data Breach Markets.
Rel. Marco Mellia, Stefano Traverso. Politecnico di Torino, Corso di laurea magistrale in Cybersecurity, 2026
Abstract
Illegal markets for credentials and personal data have been available on the dark web through anonymity services like Tor and I2P. However, large-scale law enforcement operations (e.g., Operation Cronos, BreachForums seizure) have possibly forced threat actors toward a "Clearnet frontier": messaging platforms and deep web forums providing comparable privacy level of anonymity with significantly decreased entry threshold. Effectively monitoring such an ecosystem reveals the limitations of the traditional threat intelligence gathering process. Classical web crawlers rely on hardcoded selectors and deterministic rules that fail against dynamic web pages, anti-bot protections, and the unstructured natural language typical of underground forums and chat rooms.
This thesis investigates whether autonomous Large Language Model (LLM) agents can overcome these limitations and power an automated deep web crawler capable of detecting hidden data breach markets
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Informazioni aggiuntive
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
