arrow
返回

CODE-SMASH: Source-Code Vulnerability Detection Using Siamese and Multi-Level Neural Architecture

delete2024-01-01
delete0
delete
OA
AI
S
Sungmin Han
H
Hyunkyung Nam
J
Jaesik Kang
K
Kwangsoo Kim
S
Seungjae Cho
S
Sangkyun Lee *
DOI:10.1109/ACCESS.2024.3432323delete
delete原文链接
delete分享
delete收藏
摘要

摘要

En 中文
The rapid evolution of software development, propelled by competitive demands and the continuous integration of new features, frequently leads to inadvertent security oversights. Traditional security practices, often reactive in nature, primarily focus on identifying known vulnerabilities, creating a significant shortfall in detecting emergent, zero-day threats. This paper introduces CODE-SMASH, a novel deep learning-based source code vulnerability detector that utilizes a Siamese neural network with a hierarchical architecture integrating BiGRU and attention mechanisms. Our experiments using real-world datasets, specifically the Chromium and Debian datasets, demonstrate CODE-SMASH's superiority over existing methods. It achieves significant improvements in detection performance across all key metrics, including accuracy, precision, recall, and F1-score, with average improvements of approximately 8.3%, 11.6%, 27.75%, and 17.7%, respectively, compared to the best-performing existing methods in our experiments. Moreover, CODE-SMASH shows its superior capability in handling complex and lengthy code sequences, with performance improvements for long-length code (60 to 80 lines) in F1 scores of 4.53 percentage points on the Chromium dataset and 5.62 percentage points on the Debian dataset compared to the second-best model's performance. We believe our research makes a significant contribution to the field of automated vulnerability detection by providing a high-precision solution to the growing challenges in software security. Furthermore, based on our findings, we anticipate that future research could enhance CODE-SMASH by expanding its generalizability to various programming languages and reducing computational demands to improve efficiency.
Keyword:
Source coding
Codes
Neural networks
Vectors
Computer architecture
Static analysis
Detection algorithms
Software development management
Code similarity
hierarchical attention network
Siamese neural network
source code vulnerability detection

期刊

IEEE Access 封面图
IEEE Access
IF:
3.6
论文数:
9.8W
被引数:
36.3W

机构

K
Korea University
学者数:
2.6K
论文数: 1.0K
被引数: 2.2W
引用论文

引用论文

Motion of Scandium Ions in Sc2C84 Observed by 45Sc Solution NMR通过 45Sc溶液NMR观察到的Sc2C84 中Sc离子的运动
err1996-06-06
err0
PREAI
errYoko Miyake; Shinzo Suzuki; Yasuhiko Kojima; Koichi Kikuchi; Kaoru Kobayashi; Shigeru Nagase; Masatsune Kainosho; Yohji Achiba; Yutaka Maniwa; Keith Fisher
err分享
err收藏
err分享
err收藏
Cardiovascular diseases risk prediction in patients with diabetes: Posthoc analysis from a matched case-control study in Bangladesh
err2021-02-15
err0
errOAAI
errSheikh Mohammed Shariful Islam; Shyfuddin Ahmed; Riaz Uddin; Muhammad U. Siddiqui; Mahsa Malekahmadi; Abdullah Al Mamun; Roohallah Alizadehsani; Abbas Khosravi; Saeid Nahavandi
err分享
err收藏
Production of reactive oxygen species from abraded silicates. Implications for the reactivity of the Martian soil
err2017-09-01
err0
PREAI
errEbbe N. Bak; Kaloyan Zafirov; Jonathan P. Merrison; Svend J. Knak Jensen; Per Nørnberg; Haraldur P. Gunnlaugsson; Kai Finster
err分享
err收藏
Proprioceptive Mechanism for Bioinspired Fish Swimming
err2021-06-08
err0
PREAI
errJ. Sánchez-Rodríguez; F. Celestini; C. Raufaste; M. Argentina
err分享
err收藏
Software Vulnerability Analysis and Discovery Using Deep Learning Techniques: A Survey
err2020-01-01
err43
errOAAI
errZeng, Peng; Lin, Guanjun; Pan, Lei; Tai, Yonghang; Zhang, Jun
err分享
err收藏
学者 查看更多内容