A graph-based investigation of bitcoin transactions

Abstract

The Bitcoin global cryptocurrency system has been the subject of several criminal cases. The Bitcoin network is a peer-to-peer system that has participants from all over the Internet. The Bitcoin protocol requires participating nodes to retain and update all transaction records; this ensures that all Bitcoin activities are accessible from a consistent transaction history database. This chapter describes a graph-based method for analyzing the identity clustering and currency flow properties of Bitcoin transactions. The method, which is scalable to large Bitcoin graphs, focuses on transactions relevant to criminal cases such as Mt. Gox. The analysis, which is performed on two years of Bitcoin transaction data, provides insights into the nature of anonymity provided by Bitcoin and how currency flows between selected users and communities of users.