摘要
由于评价指标间常存在关联,本文提出了一种基于关联的层次评价方法.该方法通过λ模糊测度对指标和指标集的重要程度建模,通过判断矩阵求解评价指标的Shapley值,通过Marichal熵求解指标和指标集的重要程度,通过Choquet积分自下而上计算方案的综合评价值.该方法仅需决策者提供评价指标Shapley值的评断矩阵,在实际中可操作性强.文章最后通过对B2C电子商务网站的评价验证了方法的合理性和有效性,并证明了指标间相互独立的评价问题是指标间存在关联评价问题的特例,同样可以采用本文所提出的方法加以解决.
In order to solve the judgement of the complex systems with interdependence, a new method of hierarchical judgement is peoposed. In the new method, the improtance of the index and indexes are modeled by λ, fuzzy measures; the Shapley valus of the indexes are calculated through the judgement matrix ; the improtance of the index and indexes are obtained based on the Marichal entropy and the synthetic values of the alternatives are gained by the Choquet integral from bottom to top. And finally, the models and methods are validated by evaluating for the B2C e-hussiness web sites, which also prove that the methods proposed here can also applied when the indexes are independent with each other.
出处
《情报学报》
CSSCI
北大核心
2007年第5期699-703,共5页
Journal of the China Society for Scientific and Technical Information
基金
国家自然科学基金资助项目（90510010）,教育部博士点基金资助项目（20050287026）,国防科技工业软科学基金项目（C06004）.
作者简介
章玲,女,1979年生,博士研究生,从事复杂系统评价与评价的研究.E-mail：pawlin@126.com.
周德群,男,1963年生,博士,教授,博士生导师,从事复杂系统评价与评价的研究.
