We analyze the magnitude of illicit activities in the Ethereum ecosystem. Using proprietary labeling data from the Blockchain Intelligence Group (BIG), we investigate the characteristics of a number of “malicious” Ethereum addresses. We first calculate the total number of transactions involving these addresses and the total amount of funds transferred through them, and then characterize smart contract addresses for ERC-20 tokens or DeFi applications, that the malicious addresses interact with. Finally, we apply machine learning techniques to identify additional “malicious” addresses by conducting a network clustering analysis within all Ethereum addresses from transactional relationships with the initial set of malicious addresses. Read more...
Measuring Illicit Activity in DeFi: The Case of Ethereum
- Conference Paper