From Language to Forensic Intelligence: Drug Name Recognition in the Cryptomarket Forum of Silk Road 2

Anonymised drug forums and online chat rooms constitute a relevant source of information for drug use. Content found on online forums can serve as reliable sources of information with a high number of discussions taking place on various topics. In these virtual environments, participants both seek and share drug-related information, while also sharing their own drug experiences with other users, producing linguistic traces that encode purchases, effects, and emerging practices. From a forensic perspective, such linguistic traces can serve as evidence, potentially revealing how criminal communities innovate and obscure meaning through the constant creation of new terms, acronyms, and variants. This study adopts a computational forensic semiotic perspective, in which drug-related terminology is conceptualised not merely as linguistic data, but as a dynamic system of signs whose meanings are continuously negotiated and reconfigured within socially and legally constrained environments. In this framework, drug names are not treated as stable labels but as context-dependent signs shaped by both community practices and external enforcement pressures. To assess the potential of online forums for early-warning monitoring, we developed a Drug Name Recognition (DNR) system based on a Conditional Random Field (CRF) model and applied it to forum data from the cryptomarket Silk Road 2. Extracted terms were classified into three categories: (i) new drug names absent from existing databases, (ii) variants of known drugs, and (iii) variants of emerging substances. Results of our analysis showed that our model enabled us to discover the presence of 232 new drug names compared to the presence of 106 traditional drug names, highlighting the extent of lexical innovation in online drug discourse. These findings demonstrate that internet-based linguistic traces constitute a valuable resource for detecting emerging substances prior to their formal documentation. Overall, the study underscores the importance of integrating natural language processing (NLP) in early-warning systems. It further suggests that online forums would represent promising sources for the early detection of drugs, thus suggesting that the use of an automated system could help national agencies to identify new drugs and produce linguistic evidence suitable for forensic intelligence.

Affiliations

Citations

APA
Chicago
FWB

Werner, R., François, T., & Bitzer, S. (2026). From Language to Forensic Intelligence: Drug Name Recognition in the Cryptomarket Forum of Silk Road 2. International Journal for the Semiotics of Law. Published. https://doi.org/10.1007/s11196-026-10503-z (Original work published 2026)