一句话：提出 DeepVD，一种基于图的神经网络漏洞检测模型，利用语句类型、后支配树和异常流图三类类分离特征，显著提升漏洞与良性代码的区分能力。
现有基于机器学习和深度学习的漏洞检测（VD）模型虽取得进展，但在区分漏洞代码与良性代码两类样本上仍存在明显局限。本文提出 DeepVD，一种基于图的神经网络 VD 模型，重点强化漏洞代码与良性代码之间的类分离特征。DeepVD 在不同抽象层次上利用三类类分离特征：语句类型（类似词性标注）、后支配树（刻画常规执行流）以及异常流图（刻画异常与错误处理流）。作者在包含 303 个项目、13,130 个漏洞方法的真实漏洞数据集上开展实验。结果表明，与当前最优的 ML/DL 漏洞检测方法相比，DeepVD 在精度上相对提升 13%–29.6%，召回率提升 15.6%–28.9%，F 值提升 16.4%–25.8%。消融研究进一步证实，所设计的特征与组件有助于 DeepVD 实现漏洞代码与良性代码的高类分离性。
The advances of machine learning (ML) including deep learning (DL) have enabled several approaches to implicitly learn vulnerable code patterns to automatically detect software vulnerabilities. A recent study showed that despite successes, the existing ML/DL-based vulnerability detection (VD) models are limited in the ability to distinguish between the two classes of vulnerability and benign code. We propose DeepVD, a graph-based neural network VD model that emphasizes on class-separation features between vulnerability and benign code. DeepVDleverages three types of class-separation features at different levels of abstraction: statement types (similar to Part-of-Speech tagging), Post-Dominator Tree (covering regular flows of execution), and Exception Flow Graph (covering the exception and error-handling flows). We conducted several experiments to evaluate DeepVD in a real-world vulnerability dataset of 303 projects with 13,130 vulnerable methods. Our results show that DeepVD relatively improves over the state-of-the-art ML/DL-based VD approaches 13%–29.6% in precision, 15.6%–28.9% in recall, and 16.4%–25.8% in F-score. Our ablation study confirms that our designed features and components help DeepVDachieve high class-separability for vulnerability and benign code.
@inproceedings{WangNWLZY23,
title = {DeepVD: Toward Class-Separation Features for Neural Network Vulnerability Detection},
author = {Wenbo Wang and Tien N. Nguyen and Shaohua Wang and Yi Li and Jiyuan Zhang and Aashish Yadavally},
booktitle = {ICSE},
year = {2023},
doi = {10.1109/icse48619.2023.00189}
}