Dependency Structure Modeling Framework Using Expert Survey Based Group Decision
This paper introduces a framework for modeling the dependency structure among elements of a complex system based on a group decision procedure with expert survey. There are two parties involved in each entry in the dependency matrix: the “influencing” element and the “influenced” element. After the “initial round” survey, experts’ opinions reflecting the two different perspectives of the two parties for each entry of the matrix are separately aggregated and contrasted. Entries with large gap between the opinions of the influencing and influenced elements are identified and classified as the “discussion items.” A group decision procedure for the “consensus round” that can reduce the gaps in the discussion items by communication the results of the initial round survey with supporting assessment rationales is proposed. Potential issues that can affect the effectiveness of the proposed framework are discussed.