Abstract: This paper introduces rough set theory to the AHP (analytic hierarchy process) method, and proposes a novel method of constructing AHP judgment matrix based on rough sets and information entropy. In view that the  attribute weights generated from the AHP have a certain degree of subjectivity and preference by using the expert's judgment matrices, this paper utilizes the rough set theory and information entropy method to construct an integrated objective attribute weight with data driven viewpoint to adjust the subjective judgment matrix of existing AHP method, and further proposes a novel AHP acquirement weights method by considering the subjectivity and objectivity simultaneously. Finally, this paper uses the practical food and beverage data from “Dianping.com” to verify the effectiveness and feasibility of the proposed method.

Key words: rough set theory, information entropy, analytic hierarchy process (AHP), attribute weight

摘要：

将粗糙集理论引入到AHP（analytic hierarchy process）方法中，提出一种基于粗糙集理论和信息熵的AHP判断矩阵构造方法。考虑到AHP方法主要是通过专家判断矩阵来获取属性权重，其决策结果具有一定的主观性和偏好性这一现象，利用粗糙集理论和信息熵方法来构造一种基于数据驱动的客观属性集成权重，并利用客观集成权重来修正已有AHP方法的主观判断矩阵，进而提出一种同时兼顾主观性和客观性的AHP权重获取方法。最后，利用大众点评美食推荐实际问题和数据，验证了所提出方法的有效性和可行性

关键词: 粗糙集理论, 信息熵, 层次分析法（AHP）, 属性权重