Partial-Fix Benchmark Set

Partial-Fix Benchmark Set是由慕尼黑大学和柏林洪堡大学创建的一个用于自动化程序修复的基准数据集，包含2204个基于部分修复的基准任务。该数据集通过爬取GitHub上1500个最受欢迎的开源C项目中的部分修复案例构建而成。数据集内容丰富，涵盖了多种程序修复场景，旨在帮助研究人员评估和改进基于部分修复的自动化程序修复技术。创建过程中，研究团队采用了严格的筛选标准和数据收集方法，确保数据集的质量和实用性。该数据集主要应用于自动化程序修复领域，特别是针对部分修复的研究，以解决软件开发中的实际问题。

The Partial-Fix Benchmark Set is a benchmark dataset for automated program repair, developed by the University of Munich and Humboldt-Universität zu Berlin. It contains 2,204 partial-fix-based benchmark tasks. This dataset is constructed by crawling partial-fix cases from the 1500 most popular open-source C projects hosted on GitHub. With rich content covering diverse program repair scenarios, it aims to help researchers evaluate and optimize partial-fix-based automated program repair technologies. During its development, the research team adopted strict screening criteria and data collection methodologies to ensure the dataset's quality and practical utility. This dataset is primarily utilized in the field of automated program repair, particularly for partial-fix-related research, to address real-world challenges in software development.