软件网络的软核研究与软件生长中的S模型趋近

Software-Kernel of Software Network and S-Model Approximation During Software Growth

针对软件代码研究的复杂性,研究软件类结构中最重要的核心部分,提取基于软件网络的软核,以降低软件理解与度量的难度.并通过一个以软核概念为基础的软核剥离算法进行了一定样本量上,软件软核的研究.分析结果表明:软核的层次与软件的升级与发展,即软件生长成正比关系.软件网络的软核每个层次中节点数与软件网络总节点数之比在软件生长过程中会趋近S模型.软件软核的S模型趋近现象在软件的生长过程中,可以被用来研究度量软件的层次性构成.软核的提取可以用来简化研究人员对软件的理解,聚焦软件设计、实现、测试时的重点类.

Because of the complexity in studies on software code and software network, the software-kernel of software network is defined and extracted to reduce the difficulty of grasping and measuring any software. Then a software-kernel layer stripping algorithm is developed to study the software-kernel with a certain amount of samples. The analysis results showed that the ordinal number of software-kernel layer of any software is directly proportional to the growth of that software, i.e. its upgrading evolution. The ratio of ordinal number of a kernel layer to the number of nodes of a software-kernel thus approximates to the S-model during the growth of software. Such a result is available to study the hierarchy of measuring softwares. Moreover, the extraction of software-kernel can be used to simplify the grasping of softwares and focus on the design, implementation and testing of key classes in softwares.