【Author】
Tharani, Jeyakumar Samantha; Charles, Eugene Yougarajah Andrew; Hou, Zhe; Palaniswami, Marimuthu; Muthukkumarasamy, Vallipuram
【Source】PROCEEDINGS OF THE IEEE 46TH CONFERENCE ON LOCAL COMPUTER NETWORKS (LCN 2021)
【Abstract】Blockchain is a digital technology built on three pillars: decentralization, transparency and immutability. Bitcoin and Ethereum are two prevalent Blockchain platforms, where the participants are globally connected in a peer-to-peer manner and anonymously perform trade electronically. The vast number of decentralized transactions and the pseudo-anonymity of participants open the door for scams, cyber frauds, hacks, money laundering and fraudulent transactions. It is challenging to detect such fraudulent activities using traditional auditing techniques, since they need more processing power, time and memory for complex queries to join combinations of tables. This paper proposes several algorithms to extract the transaction-related features from the Bitcoin and Ethereum networks and to represent the features as graphs. Moreover, the paper discusses how visualisation of graphs can reflect the anomalies and patterns of fraudulent activities.
【Keywords】Bitcoin; Ethereum; smart contracts; graph models; anomaly detection; blockchain
评论