【Author】
Zhang, Lejun; Li, Yuan; Guo, Ran; Wang, Guopeng; Qiu, Jing; Su, Shen; Liu, Yuan; Xu, Guangxia; Chen, Huiling; Tian, Zhihong
【Source】JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY
【Abstract】With the development of blockchain technology, smart contracts have attracted a lot of attention in recent years. They are widely used because they can reduce the cost of trust compared with traditional contracts. At the same time, they are at risk of being hacked. Therefore, the current research on smart contract vulnerability detection is particularly important. We proposed a novel smart contract reentrancy vulnerability detection model based on BiGAS. We had conducted numerous experiments, and the experimental results showed that our model (BiGAS Detection Model) has a strong vulnerability detection ability. It achieves an accuracy and F1-score of over 93% for the detection of reentrancy vulnerabilities in smart contracts. To verify that the choice of SVM is one of the reasons for improving the performance of our method, Softmax was replaced by the SVM classifier in the model. The accuracy of the model with the classifier replaced with Softmax was 89.78% and the F1-score was 89.83%. In addition, we compared our approach with advanced automated audit tools and other deep learning-based vulnerability detection methods. Compared with the existing advanced methods, the accuracy and F1-score improvement of our model ranges from 4 to 23%. The conclusion was that our method is significantly better than the existing advanced methods in detecting smart contract reentrancy vulnerabilities.
【Keywords】Smart contract; Reentrancy vulnerability; Bidirectional gated recurrent neural network; SVM
【文献类型】Article; Early Access
评论