Date

2023-2-01

Author

BEKTAŞ, Barış Cem

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

566
views

903
downloads

Cite This

Smart contracts, which are self executing contracts with the terms of the agreement between buyer and seller being directly written into lines of code, have the potential to revolutionize many industries by automating complex processes and reducing the need for intermediaries. However, the immutability of smart contracts also means that vulnerabilities cannot be easily fixed once they are deployed, making it crucial to detect and prevent vulnerabilities before deployment. In this project, we focus on the problem of vulnerability detection in smart contracts, specifically the reentrancy vulnerability, which allows an attacker to repeatedly call an external contract in a malicious manner. To address this problem, we introduce four-layer convolutional neural network (CNN) for reentrancy vulnerability scanning. We compare our method to other vulnerability scanning tools which are using machine learning approaches, including long short-term memory (LSTM) and graph neural network (GNN), and show that our method outperforms on dataset of real-world smart contracts. Our results demonstrate the effectiveness of using deep learning for vulnerability detection in smart contracts and provide a promising direction for further research in this area.

Subject Keywords

Blockchain, Smart Contracts, Deep Learning,, CNN, Vulnerability Detection

URI

https://hdl.handle.net/11511/101981

Collections

Graduate School of Informatics, Term Project