软件缺陷是影响软件生产效率和质量的关键因素，缺陷发现与修复对软件生产效率和质量有着至关重要的影响。软件变更记录了缺陷的引入与修复历史，如何从中高效且准确地吸取营养，即时地发现和定位缺陷，不断提高软件生产效率和质量，是当前软件工程领域的研究热点。本项目以软件变更为研究对象，以软件变更历史为数据基础，以即时的变更级缺陷预测与定位的融合机制为核心，系统地研究基于多维信息融合的变更级即时缺陷预测方法、基于代码自然性的变更级即时缺陷定位方法、工作量感知的变更级即时缺陷预测与定位的融合方法等关键科学问题；拟面向软件变更提出一套工作量感知的即时缺陷预测与定位的融合方案，提高缺陷修复的效率和准确率，节省缺陷修复和软件维护的成本，进而达到提高软件生产效率和质量的目的，助力我国软件产业的发展。

Software defect is the key factor that has an impact on the efficiency and quality of software production. Discovering and repairing defect has a significant impact on the efficiency and quality of software production. Software changes record the history of defect introducing and repairing. In order to discover and repair defects timely and improve the efficiency and quality of software production, how to learn knowledge from software changes effectively and efficiently is a hot research question in software engineering. This project focuses on software change as the research object, based on the software change historical data, with the just-in-time and fused defect prediction and localization as the core. This project investigates the following research questions: the change-level just-in-time defect prediction based on fusing multi-dimension features, change-level just-in-time defect localization based on software naturalness, and effect-aware change-level fused method for just-in-time defect prediction and localization. As a result, this project plans to propose a fused solution for change-level just-in-time defect prediction and localization. The aim is to improve the efficiency and accuracy of defect repair, save the effort of defect repair and software maintenance, improve the efficiency and quality of software production and finally assist the development of the software industry of our country.