Graph neural network-based transaction link prediction method for public blockchain in heterogeneous information networks

Abstract

Public blockchain has outstanding performance in transaction privacy protection because of its anonymity. The data openness brings feasibility to transaction behavior analysis. At present, the transaction data of the public chain are huge, including complex trading objects and relationships. It is difficult to extract attributes and predict transaction behavior by traditional methods. To solve these problems, we extract transaction features to construct an Ethereum transaction heterogeneous information network (HIN) and propose a graph neural network (GNN)-based transaction prediction method for public blockchains in HINs, which can divide the network into subgraphs according to connectivity and increase the accuracy of the prediction results of transaction behavior. Experiments show that the execution time consumption of the proposed transaction subgraph division method is reduced by 70.61% on average compared with that of the search method. The accuracy of the proposed behavior prediction method also improves compared with that of the traditional random walk method, with an average accuracy of 83.82%.