Signals in the Noise: Understanding User Preferences and Disclosure Patterns with AI

Conference by Raphael Hoheisel
Tuesday, November 11, 2025, from 11:45 a.m. to 1:00 p.m.
Room C-2059, Lionel-Groulx Pavilion, Université de Montréal, and on Zoom.

Registration for in-person attendance

Registration for Zoom

(After registering, a confirmation email will allow you to add this event to your calendar.)

---

Abstract:

This presentation explores the use of machine learning and large language models (LLMs) to reveal online user preferences and behavioral patterns within cybercrime forums. By analyzing user interactions and communication patterns, our studies reveal how participants navigate between implicit and explicit engagement in illicit activities, often using ambiguity to manage risk and maintain plausible deniability. We also examine the selection of private communication platforms, demonstrating that this is influenced by the context of criminal discussions and the position of users within the social network. Our findings emphasize the subtle ways in which users balance visibility, trust and security. 

Biography: 

Raphael Hoheisel

Raphael is a PhD candidate at the University of Twente in collaboration with the Dutch Police. He has a background in engineering and computer science with a focus on cyber-security. At the moment his research focuses on underground forums investigating how users disclose contact details and crime involvement.