Abstract: Information aggregation method is an important research in group evaluation. This paper proposes a method to determine the satisfaction degree of group evaluation information, and further researches a group information aggregation method. In order to further improve the fairness and accuracy, a different weight instead of a fixed one is placed on each index from the same expert in this method. Firstly, the evaluation situation and question are developed, and the satisfaction degree of the evaluation information is defined. Secondly, the methods to determine the satisfaction degree of priori information and the one of current information are presented respectively. The former is used to determine initial weights of experts, and the latter is used to modify those weights. A weight matrix of every expert can be determined in this process. Finally, the final aggregation result of group information is obtained by using these weight matrixes.

Key words: group evaluation, information aggregation, satisfaction degree of evaluation information, weights of experts, weight matrix

摘要： 信息集结方法是群体评价的重要研究内容,针对该问题本文对评价信息满意度进行测度,并基于此对群体信息的集结方法展开研究。该方法根据指标信息的变化赋予评价者不同的权重,以改变现有研究中评价者权重大多固定不变的做法,旨在使群体信息集结的结果更加公正和准确。首先对问题进行界定并给出评价信息满意度的定义;然后分别给出先验信息满意度和评价过程中的信息满意度的确定方法,利用先验信息满意度确定各评价者的初始权重,并利用评价过程中的信息满意度对其修正,从而得到各评价者的权重矩阵。最后,按照各评价者的评价信息满意度对群体评价信息进行集结。

关键词: 群体评价, 信息集结, 评价信息满意度, 评价者权重, 权重矩阵