Exploring the Evolution of Exploit-Sharing Hackers: An Unsupervised Graph Embedding Approach

This research develops a framework to track how influential hackers in dark-web forums change and grow over time, rather than just snapshotting them at one moment. It combines each hacker's activity statistics, network position, and writing style using graph-based machine learning, with a technique to compare these profiles across multiple time periods. In a case study of an English forum with over 51,000 posts, the method successfully pinpointed key hackers distributing pirated material.