一种基于模板的档案图像压缩新方法.pdfVIP

一种基于信息检索技术识别重复缺陷模式的方法 吴 倩 王千祥 （北京大学信息科学技术学院软件研究所 北京 100871） （高可信软件技术教育部重点实验室） 摘 要 缺陷模式是对常见代码缺陷的统一描述。缺陷模式库收集这些模式，并为基于缺陷模式的静态代码分析工具的研发提供支 持。当用户向缺陷模式库添加新缺陷模式时，经常发生重复提交的现象，造成库中缺陷模式的冗余。本文提出了一种基于信息检索 技术识别重复缺陷模式的方法，通过计算自然语言信息的相似度，发现语义相近的缺陷模式。实验表明，该方法能在节省用户和审 核员精力的同时，有效地减少重复缺陷模式的数量。此外，本文还对缺陷模式的表示形式进行了设计，以有利于降低重复缺陷模式 的数量。 关键词 重复，缺陷模式，相似度，自然语言处理，信息检索 中图法分类号：TP311 文献标识码：A An Approach to Detect Duplicate Defect Patterns based on Information Retrieval WU Qian, WANG Qian-xiang (Institute of Software, School of Electronics Engineering and Computer Science, Peking University, Beijing, 100871, China) (Key laboratory of High Confidence Software Technologies, Ministry of Education, China) Abstract Defect Pattern is a unified description for common code defects. By collecting these patterns, Defect Pattern Repository provides support for the research and development of Static Analysis Tools based on Defect Patterns. When new patterns are added, duplicate submission often occurs which leads to a redundancy in the repository. This paper introduces an approach to detect potential duplicates using information retrieval based on natural language processing. Evaluation indicates that the approach could effectively decrease the number of duplicates while remarkably reducing the time and effort the users or triagers spend on each pattern. This paper further researched the composition of defect patterns, which is also instrumental to reduce the number of duplicates. Key words Duplicate, Defect Pattern, Similarity, Natural Language Processing, Information Retrieval 一条缺陷模式至少包含这几个部分：名称、描述、错误 1 引言