• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.31 no.4
• /
• pp.715-729
• /
• 2021

Bitcoin is one of the cryptocurrencies, which is decentralized and transparent. However, due to its anonymity, it is currently being used for the purpose of transferring funds for illegal transactions in darknet markets. To solve this problem, clustering heuristic based on the characteristics of a Bitcoin transaction has been proposed. However, we found that the previous heuristis suffer from high false negative rates. In this study, we propose a novel heuristic for bitcoin clustering using off-chain data. Specifically, we collected and analyzed user review data from Silk Road 4 as off-chain data. As a result, 31.68% of the review data matched the actual Bitcoin transaction, and false negatives were reduced by 91.7% in the proposed method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.10a
• /
• pp.22-26
• /
• 2018

In this paper, we propose a model that classifies different transaction information by clustering and learning through similarity and transaction pattern of cryptocurrency transaction information. By using characteristics of genetic algorithms, we can get better clustering performance by eliminating unnecessary elements in clustering process. The transaction information including the clustering value is set as the training data, and the transaction information can be predicted through the classification algorithm. This can be used to automatically detect abnormal transactions from various transaction information of the cryptocurrency.