东北大学学报（自然科学版）

东北大学学报:自然科学版 ›› 2013, Vol. 34 ›› Issue (1): 40-44.DOI: -

• 论著 • 上一篇    下一篇

软件宏观拓扑结构标准结构熵和度的演化分析

徐久强，卢锁，刘铮，赵海   

  1. (东北大学信息科学与工程学院，辽宁沈阳110819)
  • 收稿日期:2012-04-20 修回日期:2012-04-20 出版日期:2013-01-15 发布日期:2013-01-26
  • 通讯作者: 徐久强
  • 作者简介:徐久强(1966-)，男，辽宁北镇人，东北大学教授；赵海(1959-)，男，辽宁沈阳人，东北大学教授，博士生导师.
  • 基金资助:
    国家自然科学基金资助项目(61101121)；教育部基本科研业务费青年教师科研启动基金资助项目(N110304003).

Evolution Analysis of Standard Structure Entropy and Degree of Software Macro-topology

XU Jiu-qiang， LU Suo， LIU Zheng， ZHAO Hai   

  1. School of Information Science ＆ Engineering， Northeastern University， Shenyang 110819， China.
  • Received:2012-04-20 Revised:2012-04-20 Online:2013-01-15 Published:2013-01-26
  • Contact: XU Jiu-qiang
  • About author:-
  • Supported by:
    -

RichHTML

1

PDF (PC)

847

摘要： 现代软件生存周期可以看做是一个软件结构的演化过程.基于大量开源软件的统计数据，分析了标准结构熵、高度值节点数量和新增节点中低度值节点所占比例的演化，然后分析了度中心化指标较高的节点数量和新增节点中该指标较高的节点数量的演化.结果表明，标准结构熵和度中心化指标较高的节点数量普遍呈下降趋势，高度值节点数量呈上升趋势，新增节点大都为低度值节点.由此得出，成功实践的软件演化过程中，软件网络和软件结构的复杂性增长速度是逐步降低的.本研究对于指导软件开发有重要的作用.

关键词: 软件网络, 软件演化, 标准结构熵, 度, 中心化

Abstract: Modern software life circle can be considered as a software architecture evolutionary process. Based on the statistical data calculated from a large number of open-source software systems， the changing of standard structure entropy， the number of high degree nodes and the ratio of new added low degree nodes were analyzed respectively. And the changes in the number of nodes with high degree centrality indicator in overall network and new added nodes were subsequently analyzed. Analysis results revealed a universal phenomenon that， the number of nodes with high standard structure entropy and high degree centrality indicator was declining， while the number of high degree nodes was increasing. Most of new added nodes were lower-degree. It could be concluded that the growth of complexity in software networks and software architecture was gradually slowing down within the process of a successful software evolution. It is significant to the guide the software development.

Key words: software networks, software evolution, standard structure entropy, degree, centralization

中图分类号: