A Synthesis of Information Imperfection Representations for Decision Aid

Abstract

Information imperfection encompasses different terms such as uncertainty, imprecision, inaccuracy or incompleteness, used in the literature to describe any “weakness” in the quality of available information. Many theories or languages such as probability theory, fuzzy set theory, possibility theory or the theory of evidence, handle these imperfections using different representation formalisms and processing modes. A review of main theories is presented emphasizing their conditions of use, some useful properties and relationships between these theories. These relationships are exploited to build an approach with high-level guidelines to assist the choice of information imperfection representation language in a decision aiding process.

Keywords:

• Information imperfections
• uncertainty modeling
• fuzzy set theory
• probability theory
• theory of evidence
• multicriteria decision aid

Notes

1 X/F refers to the set of Attributes/Criteria used in the decision model. Despite differences in the literature as per the definition of an attribute versus a criterion (e.g., Roy, 1990), they are used interchangeably in this paper.