Please use this identifier to cite or link to this item:
Title: Designing a security evaluation tool for blockchain smart contracts
Authors: Han, Xing Jie
Keywords: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
DRNTU::Engineering::Computer science and engineering::Computing methodologies::Pattern recognition
Issue Date: 2019
Abstract: Past security incidents of smart contracts on the Ethereum blockchain has proved to be disastrous - incurring losses of upwards of a few hundred million USD to date. As the attacks occurred due to semantic errors in smart contracts itself, specialized security tools which employed traditional software techniques such as symbolic analysis proved to be valuable in the detection of such vulnerable contracts; however its coverage and efficiency is limited by factors such as the depth of its search which comes at a cost of execution time. Meanwhile, the adoption of smart contracts on Ethereum has increased 176-fold since December 2015 – if these tools fail to keep up with the growth of contracts, similar incidents on a greater scale might occur in the future. In this project, we aim to contribute to the security landscape of smart contracts by proposing an efficient smart contract vulnerability detection system. We explored the approach of machine learning to vulnerability detection in smart contracts and trained a long-short term memory (LSTM) model on approximately 1.7 million contracts obtained from Google's BigQuery dataset and achieved encouraging results. We observed a detection accuracy of 99.40%, with a recall score of 89.81% in detecting vulnerable smart contracts, accompanied with significantly better performance with the model taking less than a tenth of the time required to classify a contract compared to that of a prominent symbolic tool such as Maian. In addition, a web application was developed to demonstrate the efficiency of our approach in classifying smart contracts at scale.
Schools: School of Computer Science and Engineering 
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
1.16 MBAdobe PDFView/Open

Page view(s) 50

Updated on Jul 17, 2024

Download(s) 50

Updated on Jul 17, 2024

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.