BROWSE
Detail
Journal Article
Blockchain feature extraction and behavioral profiling for illicit transaction analysis
Cited - time in 
Share 
Share
- Authors
- Mahdi Daghmehchi Firoozjaei, MinChang Kim, Hyoungshick Kim, Arash Habibi Lashkari, Ali A. Ghorbani
- Issue Date
- 2026-09
- Citation
- Forensic Science International: Digital Investigation, v.58, pp.1-12
- ISSN
- 2666-2825
- Publisher
- Elsevier
- Language
- English
- Type
- Journal Article
- Abstract
- Cryptocurrencies have become a dominant medium for financial transactions in dark web ecosystems, enabling illicit activities through pseudonymous and decentralized payment mechanisms. Understanding and monitoring such activities require systematic analysis of blockchain transaction data. Blockchain profiling provides a principled approach for characterizing user behavior, identifying suspicious patterns, and tracing illicit fund flows in adversarial environments. In this paper, we examine the goals and challenges of blockchain-based user profiling in dark web contexts, highlighting the impact of pseudonymity, obfuscation techniques, data sparsity, and adversarial adaptation. We further introduce a structured taxonomy of blockchain features designed to support comprehensive profiling of illicit activity, encompassing transactional, temporal, structural, behavioral, and obfuscation-related dimensions. To validate the relevance of the proposed taxonomy, we conduct an empirical analysis using the Elliptic Bitcoin dataset, comprising 203,769 transactions and 234,355 fund-flow edges, to extract representative profiling features and compare the behavioral characteristics of illicit and licit transactions using statistical significance testing. Our results reveal consistent and statistically supported behavioral differences in illicit and licit transaction patterns, while also identifying features with limited discriminative power. These findings highlight key limitations and provide insights for developing more robust, multi-feature blockchain profiling frameworks for dark web ecosystems.
- Keyword
- Blockchain forensics, Cryptocurrency transactions, Dark web, Behavior profiling, Illicit financial activity
- KSP Keywords
- Behavior profiling, Behavioral characteristics, Dark web, Discriminative Power, Statistical significance testing, Systematic analysis, Transaction analysis, Transaction data, User profiling, blockchain, data sparsity