A novel hybrid feature fusion model for detecting phishing scam on Ethereum using deep neural network

The development of blockchain technology has brought prosperity to the cryptocurrency market and has made the blockchain platform a hotbed of crimes. As one of the most rampant crimes, phishing scam has caused a huge economic loss to blockchain platforms and users. In order to address the threat to the financial security of blockchain, this paper proposes a model based on hybrid deep neural network to detect phishing scam accounts, namely LBPS (LSTM-FCN and BP neural network-based Phishing Scam accounts detection model), and verifies its effectiveness on Ethereum. The LBPS model provides a novel approach to analyse transaction records by adopting the BP neural network to obtain the implicit relationship between features extracted from transaction records and the LSTM-FCN neural network to capture the temporal feature from all transaction records of a target account. The experimental results demonstrate that the features selected in this paper could identify phishing scam accounts effectively. Moreover, the LBPS model performs better than the existing methods and baseline models with an F1-score of 97.86%.