基于复杂网络的软件结构度量方法综述  

Survey of the measurement of software structures based on complex networks

作　　者：孙世温[1,2] 夏承遗[1,2] 王莉[1,2] 

机构地区：[1]天津理工大学计算机与通信工程学院,天津300191 [2]天津理工大学天津市智能计算及软件新技术重点实验室,天津300191

出　　处：《智能系统学报》2011年第3期208-212,共5页CAAI Transactions on Intelligent Systems

基　　金：国家自然科学基金资助项目(60904063);天津市应用基础及前沿技术研究计划资助项目(11JCYBJC06600);中国博士后研究基金资助项目(20090460694);天津市高等学校科技发展基金资助项目(20090717;20090811;20090813);国家大学生创新性实验计划资助项目(091006007;101006019)

摘　　要：计算机软件复杂性与软件质量、开发成本和生产效率等密切相关,软件复杂性的度量和控制是计算机科学的挑战性问题之一.近年来复杂网络研究的兴起为研究软件系统结构复杂性提供了新的理论、方法和工具,该方法克服了传统软件结构度量方法侧重微观统计、缺乏全局性和整体性等缺点,构成了复杂网络与传统软件工程的交叉研究领域.对该领域的研究工作进展进行了介绍,从软件网络结构特征分析、建模以及研究成果的初步应用3个方面总结已有工作,并对今后有意义的研究方向进行了展望,如基于加权模型的软件网络建模、软件网络动态演化机制等.Software complexity is closely related to software quality, development cost, and development efficiency. Thus the measurement and control of software structural complexity is one of the most challenging problems in computer science. Recently, the research of complex networks has been providing new theories, methods, and tools for the study of software structural complexity, and overcoming some shortcomings of traditional measurement methods, such as only focusing on the microstructures and a lack of structural integrality, which results in the interdisciplinary field of complex network and traditional software engineering. In this paper, the following aspects of the research work in this new area were summarized and reviewed： topological structure analysis, network modeling, and the application of complexity controls and structural optimization. A perspective for meaningful future research emphasis was given including aspects such as network modeling based on weighted models and dynamical evolving mechanism of software systems.

关 键 词：复杂网络 软件结构复杂性 软件度量 建模 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]