Decoding Issue Dynamics: A Novel Approach for Identifying Issue Linkages in Github Projects

作     者：Yang, Wenhua He, Yingying Pan, Minxue Zhou, Yu 

作者机构：College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China State Key Laboratory for Novel Software Technology Nanjing University Nanjing China 

出 版 物：《SSRN》 

年 卷 期：2024年

核心收录：

主　　题：Adversarial machine learning 

摘      要：GitHub serves as a pivotal platform for open-source software development, where effective issue management is crucial for project success. Despite its importance, current tools in GitHub fall short in identifying interconnections among issues, a key component for efficient management. This paper addresses the challenge of detecting and classifying linkages between issues in GitHub, which is hindered by the platform s rudimentary linkage features. We introduce a novel two-stage machine learning approach that first ascertains the presence of linkages using a lightweight binary classifier and then classifies their types with a detailed multi-classifier. Our approach leverages existing issue data to enhance linkage detection, thereby facilitating better management practices by categorizing issue relations and prioritizing development efforts. This approach not only refines issue linkage identification but also optimizes computational resources by focusing detailed analysis on a filtered subset of linked issues. We demonstrate the effectiveness of our approach through comparative studies and highlight its potential for improving issue tracking in large-scale GitHub projects. © 2024, The Authors. All rights reserved.