Multi-model Smart Contract Vulnerability Detection Based on BiGRU

Shuxiao Song, Xiao Yu, Yuexuan Ma, Jiale Li, Jie Yu

Abstract: Smart contracts have been under constant attack from outside, with frequent security problems causing great economic losses to the virtual currency market, and their security research has attracted much attention in the academic community. Traditional smart contract detection methods rely heavily on expert rules, resulting in low detection precision and efficiency. This paper explores the effectiveness of deep learning methods on smart contract detection and propose a multi-model smart contract detection method, which is based on a multi-model vulnerability detection method combining Bi-directional Gated Recurrent Unit (BiGRU) and Synthetic Minority Over-sampling Technique (SMOTE) for smart contract vulnerability detection. Through a comparative study on the vulnerability detection of 10312 smart contract codes, the method can achieve an identification accuracy of 90.17% and a recall rate of 97.7%. Compared with other deep network models, the method used in this paper has superior performance in terms of recall and accuracy.