Software vulnerability availability prediction method based on knowledge graph

Abstract

The invention provides a software vulnerability availability prediction method based on a knowledge graph, and the method specifically comprises the steps: obtaining a vulnerability description text according to the data of a software vulnerability library, extracting entities, entity attributes and relationships among the entities from the vulnerability description text, constructing a software vulnerability knowledge graph, and then carrying out the expert revision of the knowledge graph; constructing a software vulnerability availability prediction model, performing graph embedding processing on the software vulnerability knowledge graph revised by the expert, constructing a training set according to the graph embedding processing, and training the software vulnerability availability prediction model; and obtaining a to-be-predicted software vulnerability description text, obtaining entities, entity attributes and relationships among the entities of the to-be-predicted software vulnerability description text, inputting the to-be-predicted software vulnerability description text into the software vulnerability availability prediction model, and outputting a corresponding software vulnerability availability prediction result by the software vulnerability availability prediction model. According to the method, the accuracy of predicating the availability of the software vulnerability of the known data set is high, and the generalization ability is very excellent.

Description

technical field [0001] The invention relates to the technical field of network security, in particular to a software vulnerability exploitability prediction method based on a knowledge graph. Background technique [0002] The existence of software vulnerabilities makes the operation of computer information systems have many security risks. In the current Internet environment, security issues are particularly important. For the numerous software vulnerabilities published every day, it is a key and difficult problem how the computer information system operation and maintenance personnel perform vulnerability maintenance and optimization for the published software vulnerabilities. Among these published software vulnerabilities, not all software vulnerabilities can be exploited to commit illegal acts. Information system operation and maintenance personnel can screen software vulnerabilities that can be exploited, and prioritize them, thereby effectively reducing network problems...

AI Technical Summary

Problems solved by technology

However, there are methods for predicting the exploitability of software vulnerabilities in the prior art. The existing prediction methods can predict the exploitability of software vulnerabilities to a certain extent, but this type of method only considers the word frequency and other information in the vulnerability description text, and does not Considering the deep semantic information between words, there are still problems such as poor prediction effect and low accuracy

Images