arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1006.2840cs.SE
收藏

基于需求工程文档的复杂性度量

A Complexity measure based on Requirement Engineering Document

Ashish Sharma, D. S. Kushwaha

摘要:

研究表明,软件质量的准确评估是软件开发中的主要问题.此外,这种估计取决于软件固有的复杂程度,即复杂性。本文试图实证证明所提出的复杂性,这是基于IEEE需求工程文件。高质量的SRS是高质量软件的先决条件。需求工程文档(SRS)是针对特定环境执行某些特定功能的特定软件产品、程序或程序集的规范。到目前为止给出的各种复杂性度量都是基于软件的代码和认知度量值的。因此,这些指标对代码开发人员没有任何影响。考虑到基于代码的方法的缺点,该方法在SDLC过程中冻结需求后立即识别软件的复杂性。建议的复杂性措施比较以及与既定的复杂性措施。最后,该趋势可以验证所提出的措施的结果。最后,基于需求的复杂性度量可以用来了解拟议软件的复杂性之前,实际实施的设计,从而节省成本和人力浪费。

英文摘要:

Research shows, that the major issue in development of quality software is precise estimation. Further this estimation depends upon the degree of intricacy inherent in the software i.e. complexity. This paper attempts to empirically demonstrate the proposed complexity which is based on IEEE Requirement Engineering document. It is said that a high quality SRS is pre requisite for high quality software. Requirement Engineering document (SRS) is a specification for a particular software product, program or set of program that performs some certain functions for a specific environment. The various complexity measure given so far are based on Code and Cognitive metrics value of software, which are code based. So these metrics provide no leverage to the developer of the code. Considering the shortcoming of code based approaches, the proposed approach identifies complexity of software immediately after freezing the requirement in SDLC process. The proposed complexity measure compares well with established complexity measures. Finally the trend can be validated with the result of proposed measure. Ultimately, Requirement based complexity measure can be used to understand the complexity of proposed software much before the actual implementation of design thus saving on cost and manpower wastage.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑