一种基于深度学习的漏洞定位系统及方法
By adding data flow and control flow edges to the abstract syntax tree to form a semantically enhanced second abstract syntax tree, and combining it with graph neural networks for fine-grained segmentation and localization, the problems of false positives and inaccurate localization in existing technologies are solved, achieving efficient and accurate vulnerability detection and localization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUAZHONG UNIV OF SCI & TECH
- Filing Date
- 2023-09-01
- Publication Date
- 2026-07-17
AI Technical Summary
Existing static application security testing tools suffer from high false positives and false negatives, and are unable to accurately locate vulnerabilities, especially in complex contexts where they struggle to detect and classify vulnerabilities.
By adding data flow edges and control flow edges to the abstract syntax tree to form a semantically enhanced second abstract syntax tree, and combining it with graph neural networks for fine-grained segmentation and localization, vulnerability detection and localization are performed using the treeLSTM vector encoding model and graph attention mechanism.
It enables accurate vulnerability location and efficient detection, reduces the workload of developers searching for vulnerabilities in the code, and improves the accuracy and efficiency of vulnerability location.
Smart Images
Figure CN117150507B_ABST