Intelligent Answer System Based on Vulnerability Knowledge Graph

Analyzing security vulnerabilities is of great significance to network protection. Although existing CVE and other databases contain rich vulnerability information, they are poorly readable and difficult to express the correlation between vulnerabilities and vulnerabilities, and between vulnerabilities and affected entities. In addition, the existing database does not support natural language retrieval, and non-professionals have certain difficulties in retrieval. Therefore, a security vulnerability ontology is designed, and then a knowledge graph is constructed; intent identification and slot filling of natural language questions are performed to complete semantic analysis; on this basis, an intelligent question answering system based on the vulnerability knowledge graph is designed and implemented. Experiments have proved that the system can simplify the user’s process of vulnerability information retrieval and query, and is convenient for mining the potential value of vulnerability database information.