面向Java程序包的代码概要自动生成技术研究

发布时间：2018-05-08 03:12

本文选题：程序理解 + 潜在语义分析　；参考：《计算机科学与探索》2017年02期

【摘要】：程序理解是从软件程序中获得抽象在程序中的功能和知识的过程,对软件维护有着重要的意义。研究表明,软件维护消耗了软件预算的50%到80%,而其中大概47%到62%的维护时间用于对软件系统的理解上。提出了一种面向Java程序的包概要方法,尝试从软件的语义层次出发,利用信息索引领域的潜在语义分析和数据挖掘领域的聚类算法对软件程序中的语义信息进行提取分析。对相似词汇的代码文件进行聚类,并从中提取话题对Java程序中的包进行刻画;对这些话题进行语义恢复,并利用MiniPar,一个英文词法分析器,来辅助生成程序中包的概要信息。实验结果表明该方法能够改进程序理解的效率。
[Abstract]:Program understanding is the process of obtaining abstract functions and knowledge from software programs, which is of great significance to software maintenance. The research shows that software maintenance consumes 50% to 80% of the software budget, and about 47% to 62% of the maintenance time is spent on understanding the software system. In this paper, a packet summary method for Java programs is proposed, which attempts to extract and analyze the semantic information in software programs by using the latent semantic analysis in the field of information indexing and clustering algorithm in the field of data mining from the point of view of the semantic level of the software. The code files of similar words are clustered and the topics are extracted to describe the packets in the Java program, and the semantic recovery of these topics is carried out. MiniPara is used as an English lexical analyzer to help generate the summary information of the packets in the program. Experimental results show that this method can improve the efficiency of program understanding.
【作者单位】：扬州大学信息工程学院;南京大学计算机软件新技术国家重点实验室;
【基金】：国家自然科学基金Nos.61402396,61472344 中国博士后面上项目No.2015M571489 软件新技术国家重点实验室开放课题项目No.KFKT2016B21 江苏省教育厅自然科学基金面上项目No.13KJB520027~~
【分类号】：TP312.2

【相似文献】