Penetration testing method, apparatus, electronic device, and computer storage medium
By constructing a domain knowledge graph using pre-trained network models and large language models, the system automatically extracts and corrects entities and their relationships in penetration testing, solving the problem of low penetration testing quality and achieving more efficient and accurate penetration testing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING HONGTENG INTELLIGENT TECH CO LTD
- Filing Date
- 2024-12-11
- Publication Date
- 2026-06-12
AI Technical Summary
Current penetration testing methods that rely on large language models suffer from poor penetration quality, particularly in terms of illusion and interpretability.
By extracting entities and their relationships from threat intelligence source data using pre-trained network models and large language models, a domain knowledge graph is constructed. Combined with the large language model, penetration testing is performed on target assets, automatically extracting and correcting entities and their relationships in attack techniques, tactics, and attack implementation processes.
It improves the quality and efficiency of penetration testing, reduces the illusions in the reasoning process of large language models, and ensures the comprehensiveness and accuracy of information on attack techniques, tactics, and the attack implementation process.
Smart Images

Figure CN122197018A_ABST