Toward Detecting Illegal Transactions on Bitcoin Using Machine-Learning Methods

Abstract

As an emergent electronic payment system, Bitcoin has attracted attention for its desirable features such as disintermediation, decentralization, and tamper-proof recording of data. The Bitcoin network also employs public key cryptography to prevent the disclosure of information related to participating users. Although the public key cryptography ensures the privacy and hides the true identity of users in the Bitcoin network, it has recently been abused for illegal activities that have tarnished the charm of this novel technology. Detecting the illegal transactions associated with illicit activities in Bitcoin is therefore imperative. This paper proposes a machine-learning based approach that classifies Bitcoin transactions as illegal or legal. The detected illegal transactions can be excluded from the subsequent block, promoting user acceptance and adoption of the Bitcoin technology.