Skip to main navigation Skip to search Skip to main content

Software Bug Prediction based on Complex Network Considering Control Flow

  • Zhanyi Hou
  • , Ling Lin Gong
  • , Minghao Yang*
  • , Yizhuo Zhang
  • , Shunkun Yang
  • *Corresponding author for this work
  • Beihang University
  • State Grid Corporation of China

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The prediction for software bug number provides vital guidance to the quality management and software testing. In this paper, a novel software bug number prediction method was proposed based on complex network considering control flow. Firstly, for each release of software, we constructed the Call Graph (CG), and for each release, Control Flow Graph (CFG) of every function were constructed. Then the CG Metrics (CGM) and CFG Metrics (CFGM) for each version were calculated with indicators from complex-network science. Finally, the results were sent to Panel Data Model (PDM) to perform the prediction on bugs fixed number. The experimental result showed that our method outperformed other prediction methods by 9.35% to 16.85%, and introducing CFGM reduced MAE by 5.1% to 27.8% than barely use CGM. The prediction of fixed bugs could indicate the software quality, and assist the quality control of software engineering.

Original languageEnglish
Title of host publicationProceedings - 2022 IEEE 22nd International Conference on Software Quality, Reliability and Security Companion, QRS-C 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages246-254
Number of pages9
ISBN (Electronic)9798350319910
DOIs
StatePublished - 2022
Event22nd IEEE International Conference on Software Quality, Reliability and Security Companion, QRS-C 2022 - Virtual, Online, China
Duration: 5 Dec 20229 Dec 2022

Publication series

NameProceedings - 2022 IEEE 22nd International Conference on Software Quality, Reliability and Security Companion, QRS-C 2022

Conference

Conference22nd IEEE International Conference on Software Quality, Reliability and Security Companion, QRS-C 2022
Country/TerritoryChina
CityVirtual, Online
Period5/12/229/12/22

Keywords

  • Bug Prediction
  • Complex Network
  • Control Flow Graph
  • Panel Data Model

Fingerprint

Dive into the research topics of 'Software Bug Prediction based on Complex Network Considering Control Flow'. Together they form a unique fingerprint.

Cite this