基于Rough Set理论的判断矩阵构造方法

群决策支持系统（GDSS）是利用信息技术，将多个决策者个人对策该领域的理解及个人判断能力结合在一起，利用结构化决策分析技术，对半结构化和非线结构化问题进行求解，层次分析法是对定性问题作定量分析的一种简便而有效的方法，提出了一GDSS环境下判断矩阵构造的新方法；运用业集理论；利用管理信息系统运行所积累的大量决策数据来分析各决策因素，决策者对决策目标的影响，并建立自学习系统，从而构造具有完全一致性的综合判断矩阵，该方法适用于重复群决策问题。

Constructed methods of judgment matrix based on rough set theory

In the group decision support system(GDSS), personal judgement and comprehension of many decision-makers to a domain are combined by communication and computer techniques to solve semi-structure and non-structure problems using the decision analysis techniques of solving structure problems. Analytic Hierarchy Process(AHP) is a convenient and efficient method which analyzes quantificationally qualitative problems. Rough Set theory is often used to find the dependency between two attributes. From this, we will propose a new method in this paper based on rough set to construct the judgment matrices using the decision data in MIS. In this paper, the dependency of attribute is defined and the method of calculating dependency is also pointed out. Meanwhile, the construction of personal judgment matrices, decision-makers' relative importance and construction of combined judgment matrix are introduced. Furthermore, a constructed algorithm of judgment matrix, CAJM, is proposed. At last, this paper provides an application example to testify CAJM. The method is very efficient to re-decision especially.