群体测试环境下高影响力缺陷报告的智能筛选、分派与辅助修复方法研究

在软件开发与维护过程中，缺陷修复是至关重要的一个环节。快速准确地为缺陷分派合适的开发人员进行修复可以有效提高缺陷修复效率。已有的工作通常根据缺陷报告的文本和相关元数据信息来分析开发人员的特点，忽略了缺陷严重性、人员参与度和工作负荷对承担任务的影响。进入到群体测试时代，互联网工人在技能和责任上的不确定性使得缺陷报告具有数量大、噪音多等特性。这些因素会导致传统研究工作给出不合理的缺陷分派，从而严重影响缺陷修复的效率。本项目拟对群体测试环境下高影响力缺陷报告的智能处理进行研究，具体研究内容如下：1) 如何建立抗噪音容忍数据分布不平衡的缺陷报告影响力识别模型；2) 如何分析缺陷修复任务的元数据信息、人员参与度和工作负荷度对高影响力缺陷修复任务进行合理分派；3) 如何利用海量历史信息来辅助开发人员修复高影响力缺陷。本项目结合已有的研究基础拟定了切实可行的研究方案，为其实际应用提供相关的理论方法指导。

Bug Fixing is an important process in software development and maintenance. The efficiency of bug repair can be effectively improved by quick and accurate triaging of the appropriate developer for the bug. Existing researches mainly focus on analyzing the text and metadata in bug reports to characterize the relationships between developers and bug reports, while the importance of bugs, the active level and work load level of developers tend to be ignored. Entering the era of collective intelligence testing, the uncertainty of skills and responsibilities of Internet workers makes bug reports have large-scale and high noise characteristics. These factors will lead to unreasonable bug triage in existing researches, which will seriously affect the efficiency of bug repair. This project intends to study the intelligent processing of high-impact bug reports in the collective intelligence testing. To be specific, this project focuses on the following aspects: 1) How to establish a high-impact bug reports identification model for the noise-tolerant and data imbalance distribution; 2) How to analyze the metadata information, personnel participation and workload of the defect repair task to reasonably triage high-impact bug repair tasks; 3) How to use massive historical information from the third party website to assist developers in repairing high-impact bugs. This project has developed a practical research plan based on the existing research foundation, and provides relevant theoretical methods and guidance for related practical application.