【摘要】 针对语言的非平衡性、信息的不确定性以及决策者的风险偏好和损失厌恶心理等特点，本文提出了一种基于非平衡语言的扩展序数偏好排序法（Technique for Order Preference by Similarity to Ideal Solution,TOPSIS）。首先，采用非平衡语言刻画决策信息，有效降低信息的损失，因此，刻画决策信息既可以使用已有的基于平衡语言这一假设的语言变量，又可以使用非平衡语言或假设以非平衡语言为基础的语言变量；其次，在使用非平衡语言刻画决策信息的环境下，采用前景理论考虑决策者的心理行为特征；再次，使用D数刻画不确定信息；再将传统的序数偏好排序法（TOPSIS）扩展到用非平衡语言表达决策信息的领域，并对其决策矩阵加以改进，形成扩展的TOPSIS多属性决策方法；将提出的方法用于医生推荐中。结果表明，具有一定的可行性和有效性。更多还原

【Abstract】 Aiming at the characteristics of linguistic imbalance,information uncertainty,risk preference and loss aversion of decision makers,this paper proposes an extended TOPSIS(Technique for Order Preference by Similarity to Ideal Solution)ordinal preference ranking method based on unbalanced linguistic term sets.First,the use of unbalanced linguistic term sets to portray decision-making information can effectively reduce the loss of information.Therefore,the portrayal of decision-making information can not only use existing language variables developed based on the assumption of balanced linguistic term sets,but also use new unbalanced linguistic term sets and balanced linguistic term sets;Secondly,in the context of unbalanced linguistic term sets describing decision-making information,prospect theory is used to consider the psychological and behavioral characteristics of decision-makers;again,the D number is used to describe uncertain information;then,the traditional ordinal preference ranking method(TOPSIS)is extended go to the field of expressing decision information in unbalanced linguistic term sets,and improve its decision matrix to form an extended TOPSIS multiple attribute decision making method;finally,the proposed method is used in doctor recommendation.The results show the feasibility of the proposed method,effectiveness and practicality.更多还原