VulKnow: Enhancing Vulnerability Detection With Structured Knowledge and Large Language Models
IEEE Internet of Things Journal · 2026-08 · 13 (16) · 36005-36020
DOI: 10.1109/jiot.2026.3697600 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Internet of Things Journal, 2026-08. Verfasst von Guixiang Liao und 4 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben. Gefördert von: National Key Research and Development Program of China, Guizhou Provincial Department of Science and Technology 2025 Targeted Support for Scientific Research Projects of Guizhou Provincial Laboratories, 2024 Xizang Autonomous Region Central Guided Local Science and Technology Development Fund Project.
Abstract (Originalsprache)
The rapid proliferation of Internet of Things (IoT) systems has led to increasingly heterogeneous, resource-constrained, and security-sensitive software deployments. In this context, detecting vulnerabilities in embedded and system-level IoT code has become particularly critical, as even a single exploitable flaw may compromise entire networks or devices. Vulnerability detection in real-world software continues to pose significant challenges due to the intricate nature of program semantics, the diversity of vulnerability patterns, and the limited explainability offered by current machine learning models. Although large language models (LLMs) have exhibited remarkable capabilities in comprehending and reasoning about source code, their effectiveness is frequently compromised by inadequate domain knowledge and instances of hallucinated outputs. To address these limitations, we propose VulKnow, a retrieval-augmented framework for vulnerability detection that harnesses structured vulnerability knowledge alongside multistage prompt-based reasoning. Specifically, VulKnow first constructs a structured knowledge base derived from authoritative sources such as CWE/CVE reports and standard library specifications. Subsequently, it conducts context-aware knowledge retrieval and integrates the retrieved items into an LLM-based initial filtering module. To ensure high-confidence and verifiable results, we design a multistage reasoning pipeline that progressively validates vulnerabilities through semantic prompts and consistency checks. Experiments conducted on multiple real-world datasets (Linux, Qemu, and Big-Vul) demonstrate that VulKnow significantly enhances precision, recall, and explainability compared to existing static analysis tools as well as LLM-only baselines. This positions VulKnow as a reliable solution for practical vulnerability auditing scenarios.
Förderung
- National Key Research and Development Program of China
- Guizhou Provincial Department of Science and Technology 2025 Targeted Support for Scientific Research Projects of Guizhou Provincial Laboratories
- 2024 Xizang Autonomous Region Central Guided Local Science and Technology Development Fund Project
- Basic Research Program for Natural Science of Guizhou Province
Zitieren
APA
Liao, G., Chen, Y., Ke, W., Wu, H., & Dong, Z. (2026). VulKnow: Enhancing Vulnerability Detection With Structured Knowledge and Large Language Models. IEEE Internet of Things Journal, 13(16), 36005-36020. https://doi.org/10.1109/jiot.2026.3697600
Vancouver
Liao G, Chen Y, Ke W, Wu H, Dong Z. VulKnow: Enhancing Vulnerability Detection With Structured Knowledge and Large Language Models. IEEE Internet of Things Journal. 2026;13(16):36005-36020. doi:10.1109/jiot.2026.3697600
BibTeX
@article{Liao2026,
author = {Liao, Guixiang and Chen, Yanli and Ke, Wei and Wu, Hanzhou and Dong, Zhicheng},
title = {VulKnow: Enhancing Vulnerability Detection With Structured Knowledge and Large Language Models},
journal = {IEEE Internet of Things Journal},
year = {2026},
volume = {13},
number = {16},
pages = {36005--36020},
doi = {10.1109/jiot.2026.3697600}
}
EUVIMED