基于相关主题模型的程序网络自动构建与分析

doi:10.3969/j.issn.0372-2112.2017.05.004

电子学报 ›› 2017, Vol. 45 ›› Issue (5): 1052-1056.DOI: 10.3969/j.issn.0372-2112.2017.05.004

基于相关主题模型的程序网络自动构建与分析

孙小兵^1,2, 刘湘月¹, 李斌^1,2, 张伟佳¹

1. 扬州大学信息工程学院, 江苏扬州 225127;
2. 南京大学计算机软件新技术国家重点实验室, 江苏南京 210023

收稿日期:2015-12-25 修回日期:2016-05-25 出版日期:2017-05-25
- 通讯作者:
- 孙小兵
- 作者简介:
- 刘湘月 女,1990年出生,江苏盐城人,硕士研究生,研究方向为程序理解.E-mail:495296335@qq.com
- 基金资助:
- 国家自然科学基金 (No.61402396，No.61472344）; 江苏省教育厅自然科学基金 (No.13KJB520027）; 南京大学计算机软件新技术国家重点实验室开放基金 (No.KFKT2016B21）; 江苏青蓝工程

On Automatic Construction and Analysis of Program Network via Relational Topic Model

SUN Xiao-bing^1,2, LIU Xiang-yue¹, LI Bin^1,2, ZHANG Wei-jia¹

1. School of Information Engineering, Yangzhou University, Yangzhou, Jiangsu 225127, China;
2. State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, Jiangsu 210023, China

Received:2015-12-25 Revised:2016-05-25 Online:2017-05-25 Published:2017-05-25

摘要/Abstract

摘要：

程序理解的目的在于获得足够的软件系统信息，以适用于人理解的形式展现出来，辅助开发人员对软件的理解.本文通过使用相关主题模型，为软件系统类层次的代码文件建立程序网络，并可视化展示整个软件系统的相关结构和功能，辅助开发者理解整个程序代码.该技术综合考虑了软件代码中的结构性信息和内容性信息，所建立的程序网络可帮助开发者更好的理解程序的语法依赖关系和语义功能相关关系.实验验证了建立的程序网络具有较好的准确性以及可以为指定的类推荐相关类.

关键词: 程序理解, 相关主题模型, 程序网络

Abstract:

Program comprehension aims to obtain enough information in the software system to promote the comprehension of the target software.This paper proposes a novel technique,which uses relational topic model (RTM) to model code (class-level) documents in the software system into a program network.Then,the program network is visualized to help developers understand the whole software.The advantage of RTM is that it takes into account both the structural and textual information in the software system,which enables developers to fully understand the syntax dependence and semantic functional relationship in the program.The empirical results show that the program network is more accurate to model the relation among different classes,moreover,it is able to recommend relevant classes for a given class to understand a local part in the program.

Key words: program comprehension, relational topic model, program network

中图分类号:

TP311

孙小兵, 刘湘月, 李斌, 张伟佳. 基于相关主题模型的程序网络自动构建与分析[J]. 电子学报, 2017, 45(5): 1052-1056.

SUN Xiao-bing, LIU Xiang-yue, LI Bin, ZHANG Wei-jia. On Automatic Construction and Analysis of Program Network via Relational Topic Model[J]. Acta Electronica Sinica, 2017, 45(5): 1052-1056.

参考文献

[1] Rajlich V.Software evolution and maintenance[A],Proceedings of the on Future of Software Engineering[C].USA:ACM,2014.133-144.
[2] Obrien M P.Software comprehension:A review and research direction[R].Tech.rep.(2003).
[3] Rajlich V,Wilde N.The role of concepts in program comprehension[A].Proceedings of the 10th International Workshop on Program Comprehension[C].USA:IEEE,2002.271-278.
[4] Chang J,Blei D.Relational topic models for document networks[A].Proceedings of the Artificial Intelligence and Statistics[C].USA:MIT Press,2009.81-88.
[5] Bavota G,Oliveto R,Gethers M,Poshyvanyk D,De Lucia A.Methodbook:Recommending move method refactorings via relational topic models[J].IEEE Transactions on Software Engineering,July 2014,40(7):671-694.
[6] Gethers M,Poshyvanyk D.Using relational topic models to capture coupling among classes in object-oriented software systems[A].Proceedings of 26th IEEE International Conference on Software Maintenance[C].USA:IEEE,2010.1-10
[7] Liu X,Sun X,Li B.PFN:A novel program feature network for program comprehension[A].Proceedings of the 13th IEEE/ACIS International Conference on Computer and Information Science[C].USA:IEEE,2014.349-354.
[8] Sun X,Liu X,Hu J,Zhu J.Empirical studies on the nlp techniques for source code data preprocessing[A].Proceedings of the 2014 International Workshop on Evidential Assessment of Software Technologies[C].USA:ACM,2014.32-39.
[9] Abdeen H,Sahraoui H A,Shata O,Anquetil N,Ducasse S.Towards automatically improving package structure while respecting original design decisions[A].Proceedings of the Working Conference on Reverse Engineering[C].USA:IEEE,2013.212-221.
[10] Sun X,Li B,Tao C,Zhang S.HSM-based change impact analysis of object-oriented java programs[J].Chinese Journal of Electronics,Apr.2011,20(2):247-251.
[11] Van Rijsbergen C J.Information Retrieval[M].Butterworth,London,1979.
[12] Sun X,Li B,Tao C,Wen W,Zhang S.Analyzing impact rules of different change types to support change impact analysis[J].International Journal of Software Engineering and Knowledge Engineering,2013,23(3):259-288.
[13] Harman M,Binkley D,Gallagher K,Gold N,Krinke J.Dependence clusters in source code[J].ACM Transactions on Programming Languages and Systems.2009,31(1):1-33.
[14] Thomas S W,Adams B,Hassan A E,Blostein D.Modeling the evolution of topics in source code histories[A].Proceedings of the Working Conference on Mining Software Repositories[C].USA:ACM,2011.173-182.
[15] Liu X,Sun X,Li B.Top-down program comprehension with multilayer clustering based on LDA[A].Proceedings of the International Workshop on Evidential Assessment of Software Technologies[C].USA:IEEE,2014.56-59.

基于相关主题模型的程序网络自动构建与分析

On Automatic Construction and Analysis of Program Network via Relational Topic Model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 5

编辑推荐

Metrics

[1]	潘伟丰, 宋贝贝, 胡博, 李兵, 姜波. 基于软件网络加权k-核分析的关键类识别方法[J]. 电子学报, 2018, 46(5): 1071-1077.
[2]	李文杰, 姜淑娟, 钱俊彦, 王兴亚, 鞠小林. 基于对象引用关系的Java程序内存行为分析方法[J]. 电子学报, 2015, 43(7): 1336-1343.
[3]	姜淑娟, 鞠小林, 王兴亚, 李海洋, 张艳梅, 刘颖祺. 基于UIO序列的类重要性度量[J]. 电子学报, 2015, 43(10): 2062-2068.
[4]	潘伟丰, 李兵, 马于涛, 姜波. 基于加权PageRank算法的关键包识别方法[J]. 电子学报, 2014, 42(11): 2174-2183.
[5]	赖冠辉, 王晓博, 刘超. 代码与文档间关联关系的提取方法研究和改进[J]. 电子学报, 2009, 37(S1): 22-30.