一种改进的类比估算方法及案例研究  

AN IMPROVED ANALOGY-BASED ESTIMATION METHOD AND ITS CASE STUDY

作　　者：李效云 杨达[1] 杨叶[1] 

机构地区：[1]中国科学院软件研究所互联网软件技术实验室,北京100190 [2]中国科学院研究生院,北京100049

出　　处：《计算机应用与软件》2011年第7期5-9,37,共6页Computer Applications and Software

基　　金：国家自然科学基金项目(90718042;60873072;60903050);国家高技术研究发展计划项目(2007AA010303)

摘　　要：在软件成本估算中类比估算所要解决的主要问题之一是如何定义合适的相似度,有必要充分利用模型信息提高估算精度。提出一种改进的类比估算方法,定义模型属性信息空间上由第一类曲线积分计算的弧长为相似度。在此基础上开发了改进的类比估算工具ImpAnalogy,并且进行了案例研究。结果表明改进后的类比估算方法有效地提高了估算精度。In software cost estimation,one of the main problems for analogy-based method to be resolved is how to define a suitable similarity.There is the necessity to make full use of model information to improve the estimation accuracy.In this paper,we proposed an improved method of analogy-based estimation,which defines the length of arc computed by line integrals of the first type in information space of model attribute as the similarity.Based on that an improved analogy-based estimation tool ImpAnalogy was developed and its case study was conducted.The result demonstrates that the improved analogy-based estimation method effectually raises the accuracy of the estimation.

关 键 词：软件成本估算 类比估算 相似度函数 第一类曲线积分 ImpAnalogy 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]