Method for identifying key classes in software system based on graph neural network

Abstract

The invention discloses a method for identifying key classes in a software system based on a graph neural network, which comprises the following steps of: (1) analyzing a software source code file byusing an open source tool, extracting various association relationships among classes, and constructing a weighted software network model among the classes according to the relationships; (2) learningnodes in the weighted software network by using a network embedding method to obtain embedded vectors of the nodes; (3) constructing a sorting model based on a graph neural network; and (4) trainingthe model by using pairwise sorting loss to realize importance sorting of nodes in the network. The invention has the advantages that under the condition that the recognition precision is ensured, a large software system can be analyzed within extremely low time complexity, and potential key classes can be recognized from the large software system. Developers understand a software structure and improve the stability of a software system.

Description

technical field [0001] The invention relates to the technical field of software engineering software network analysis, in particular to a method for identifying key classes in a software system based on a graph neural network. Background technique [0002] As the scale of the network continues to expand, the complexity of the software system continues to increase, and defects in the software are inevitable. It is becoming more and more difficult to find these defects. Once these defects are exposed during the use of software, it often affects not only the module itself where the defect is located, but the entire software system may collapse. Some studies have pointed out that defects in software systems are similar to the propagation model of diseases, and will spread to other modules without defects along with the dependencies between various modules in the software system, such as method calls and parameter transfers, resulting in no defects. The module of the last also h...

AI Technical Summary

Problems solved by technology

It is becoming more and more difficult to find these defects. Once these defects are exposed in the process of using the software, it often affects not only the module itself where the defect is located, but the entire software system may collapse.

When the above methods calculate the importance of nodes, most of them are based on the local information of nodes, and some methods based on global information have high computational complexity in real large-scale networks.

In addition, the above methods are all carried out on the basis of manually constructing metrics, and the graph neural network has not been used to automatically learn the global structural information of the network and apply the learning results to the key class identification research of software systems.

Images