Leveraging historical project information to understand and support software maintenance and change activities

利用历史项目信息来理解和支持软件维护和变更活动

• 批准号: 327521-2006
• 负责人: Hassan, Ahmed
• 金额: $ 1.32万
• 依托单位: Queen's University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2007
• 资助国家: 加拿大
• 起止时间: 2007-01-01 至 2008-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=364287
• 关键词: Leveraging; historical; project; information; understand

Software repositories (such as source control systems, archived communications between project personnel, and defect tracking systems) contain a wealth of historical information about the progress of software projects. For example, source control systems store changes to the code as development progresses and defect tracking systems follow the resolution of software defects. Practitioners actively use such information in ad-hoc manners to effectively manage their projects and develop high quality code; yet few software engineering researchers investigate enriching their current research methods and techniques with historical information.     The long term goal of this research is to systematically uncover the potential and empirically explore the benefits of information stored in software repositories. This proposal will mine such historical information to understand and support software change activities. The recovered information will be integrated into tools, development environments and processes used by software practitioners to build high quality systems that are delivered within budget and on schedule.     In particular, this research will create models to understand, predict and explain code change patterns (such as refactorings) in large projects over the short (inter-release) and long (across releases) term. Since practitioners require immediate and timely access to the most appropriate and knowledgeable person related to a piece of code or feature, this research will develop techniques and tools to support practitioners in effectively locating code experts (i.e. other developers who are most familiar with particular code sections). This proposal will as well explore improving the learning experience of junior students by creating automated code tutors to assist students in producing bug free code and to inform instructors about any difficulties students are having in applying recently taught concepts. All the proposed research will be validated through large empirical studies.

软件存储库（如源代码控制系统、项目人员之间的存档通信和缺陷跟踪系统）包含大量关于软件项目进度的历史信息。例如，源代码控制系统在开发过程中存储对代码的更改，缺陷跟踪系统遵循软件缺陷的解决方案。从业者积极使用这些信息在特设的方式来有效地管理他们的项目和开发高质量的代码，但很少有软件工程研究人员调查丰富他们目前的研究方法和技术与历史信息。 本研究的长期目标是系统地揭示存储在软件存储库中的信息的潜力和经验探索的好处。本提案将挖掘此类历史信息，以理解和支持软件变更活动。恢复的信息将被整合到软件从业人员使用的工具、开发环境和流程中，以建立在预算范围内按时交付的高质量系统。 特别是，这项研究将创建模型来理解，预测和解释大型项目中短期（跨版本）和长期（跨版本）的代码更改模式（如重构）。由于从业者需要立即和及时地访问与一段代码或功能相关的最合适和最有知识的人，本研究将开发技术和工具，以支持从业者有效地定位代码专家（即最熟悉特定代码部分的其他开发人员）。该提案还将探索通过创建自动代码导师来帮助学生生成无错误代码，并告知教师学生在应用最近教授的概念时遇到的任何困难，从而改善低年级学生的学习体验。所有提出的研究都将通过大型实证研究进行验证。