Exploring the characteristics of issue-related behaviors in GitHub using visualization techniques

Zhifang Liao, Dayu He, Zhijie Chen, Xiaoping Fan*, Yan Zhang, Shengzong Liu*

*Corresponding author for this work

    Research output: Contribution to journalArticle

    47 Downloads (Pure)

    Abstract

    Feedback from software users, such as bug reports, is vital in the management of software projects. In GitHub, the feedback is typically expressed as new issues. Through filing issue reports, users may help identify and fix bugs, document software code, and enhance software quality via feature requests. In this paper, we aim at investigating some characteristics of issues to facilitate issue management and software management. We investigate the important degrees of behaviors that are related to issues in popular projects to assess the importance of issues in GitHub and analyze the effectiveness of issue labeling for issue handling. Then, we explore the patterns of issue commits over time in popular projects based on visual analysis and obtain the following results: we find that the behaviors that are related to issues play important roles in the GitHub. We also find that the time distribution of issue commits follows a three-period development model, which approximately corresponds to the project life cycle. These results may provide a new knowledge about issues that can help managers manage and allocate project resources more effectively and even reduce software failures.
    Original languageEnglish
    Pages (from-to)24003-24015
    Number of pages13
    JournalIEEE Access
    Volume6
    Early online date28 Feb 2018
    DOIs
    Publication statusPublished - 2018

      Fingerprint

    Keywords

    • open-source software community
    • project development model,
    • visual analysis
    • issue commit
    • software management

    Cite this