Practical Automated Software Bug Fixing via Syntactic and Semantic Analyses

这项提议旨在推动自动化软件错误修复的实际采用，这是最近被行业采用的，例如Facebook。它将产生新的方法，使用挖掘软件库、程序分析和人工指导的搜索来帮助自动修复规模和准确。预期的结果包括一个公开可用的自动化错误修复框架。该项目将帮助软件行业向用户交付具有更高可靠性和安全性的高质量软件，并通过自动生成反馈来提高学生学习编程的教育质量。

This proposal aims to advance the practical adoption of automated software bug repair, which has recently been adopted by industry, e.g., Facebook. It will produce novel methods that use mining software repositories, program analysis, and human-guided search to help automated repair to scale and be accurate. Expected outcomes include a publicly available automated bug repair framework. This project will help the software industry deliver to users high quality software with improved reliability and safety, and increase education quality for students learning to code via automated feedback generation.