Graph neural network-based vulnerability identification and prediction method and system, computer equipment and storage medium

Abstract

The invention discloses a graph neural network-based vulnerability identification and prediction method and system, computer equipment and a storage medium. The method comprises the following steps: constructing a vulnerability data set; dividing the vulnerability data set into a training set and a test set; expressing a vulnerability file code graph; vulnerability feature extraction; and constructing a predictor, and predicting vulnerabilities in the code file by utilizing the predictor. The system is used for realizing the process of the method, and the computer equipment and the storage medium can realize the process of the method by executing computer programs. According to the method, the grammar and semantic information of the vulnerability codes can be better utilized, the relationship between the vulnerability codes and the context is fully mined, one type of vulnerability is effectively identified, the universality and universality are higher, the link of manually formulatingvulnerability indexes in actual code auditing can be replaced, the actual use cost is lower, the application field is wider, and the precision is higher.

Description

technical field [0001] The invention belongs to the field of software engineering, and in particular relates to a method, system, computer equipment and storage medium for identifying and predicting vulnerabilities based on a graph neural network. Background technique [0002] Vulnerability identification and prediction is an important part of the software maintenance process. In recent years, with the expansion of the scale and complexity of software projects, a large number of vulnerabilities have appeared in the software development process. How to identify and predict vulnerabilities accurately and efficiently has become a very challenging task. In the previous work, the characteristics or patterns manually formulated by human experts were used as input by machine learning algorithms to detect vulnerabilities. However, some vulnerability metrics (such as code size, cyclomatic complexity, etc.) are manually defined by experts, which is too costly and expensive. Higher su...

AI Technical Summary

Problems solved by technology

There are also some works that analyze vulnerable code at a finer granularity (function level) by introducing graphs into code representations, such as the document "Vulnerabilityextrapolation: assisted discovery of vulnerabilities using machine learning" by integrating abstract syntax trees, control flow graphs, and program dependencies. Graphs form code property graphs to represent source codes, but do not propose a complete recognition prediction model

Images