A (Semi-)Supervised NLP approach for Identifying \& Linking Vendor Migrants \& Aliases on Darknet Markets

Anonymous

Abstract: The anonymity on the Darknet allows vendors to stay undetected by using multiple vendor aliases or frequently migrating between different markets. Consequently, illegal markets and their connections are challenging to uncover on the Darknet. To identify relationships between illegal markets and their vendors, we propose VendorLink, an NLP-based approach that examines writing patterns to verify, identify, and link unique vendor accounts across the advertisements (ads) on seven public Darknet markets. In contrast to the existing vendor verification literature, VendorLink utilizes the strengths of supervised learning, semi-supervised learning, and knowledge transfer to verify and identify migrating vendors and their potential aliases with state-of-the-art (SOTA) performance on both existing and emerging low-resource (LR) Darknet markets. As a result, our approach can better aid law enforcement agencies (LEA) make more informed decisions by offloading labour and helping them effectively utilize manual resources.

Paper Type: long