摘要
针对理想软件产品方案选择研究中的不足,文章将单个顾客用多粒度语言来描述的软件产品功能需求属性重要度进行基本语言术语(BLTS)转换,在此基础上引进诱导有序加权平均(IOWA)算子来确定某类顾客对于软件产品所有的产品功能需求属性重要度的评价值;根据各软件产品方案的工程属性特征值运用数据包络分析(DEA)方法来确定各软件产品方案的排名权值;运用效用值法将基于不同需求因素的软件产品方案排名决策信息进行转换,再采用互补判断矩阵中的排序公式来选择最贴近某类顾客需求的理想软件产品方案。最后给出了算法释例。
First,multi-granularity is transferred to basic linguistic term set when each customer gives Importance of software products＇ requiring function attributes.With IOWA operator,importance of products＇ requiring function attributes are determined for a kind of customers.Weight of software product for the order is calculated considering of their engineering attributes＇ value by DEA.Based on the translation formulas of utility values and ranking formulas of complementary judgment matrix for information integrated,an ideal software product solution is selected,which is most fitting customers＇ requirement.At last,effectiveness of this method is confirmed with an example.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第15期7-11,共5页
Computer Engineering and Applications
基金
航空科学基金项目(编号:03J53074)
关键词
多粒度语言
IOWA算子
数据包络分析
软件产品
multi-granularity, IOWA operator, analysis of data envelopment, software product
