Blockchain Scam Detection: State-of-the-Art, Challenges, and Future Directions

With the rapid development of blockchain platforms, such as Ethereum and Hyperledger Fabric, blockchain technology has been widely applied in various domains. However, various scams exist in the cryptocurrency transactions on the blockchain platforms, which has seriously obstructed the development of blockchain. Therefore, many researchers have studied the detection methods for blockchain scams. On the basis of introducing the mainstream types of scams, including Ponzi scheme, Phishing scam, Honeypot, and Pump and dump, this paper provides a thorough survey on the detection methods for these scams, in which 48 studies are investigated. The detection methods are categorized into the analysis-based methods and the machine learning-based methods in terms of the adopted techniques, and are summarized from multiple aspects, including the type of dataset, the extracted feature, the constructed model, etc. Finally, this paper discusses the challenges and potential future research directions in blockchain scam detection.