Loading…

DSmT-Based Group DEMATEL Method with Reaching Consensus

The decision-making trial and evaluation laboratory (DEMATEL) method employs expert assessments expressed by crisp values to construct a group initial direct - relation (IDR) matrix. However, it tends to be a low-precision expression, especially in complex practical problems. Although significant ef...

Full description

Saved in:
Bibliographic Details
Published in:Group decision and negotiation 2019-12, Vol.28 (6), p.1201-1230
Main Authors: Du, Yuan-Wei, Zhou, Wen
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The decision-making trial and evaluation laboratory (DEMATEL) method employs expert assessments expressed by crisp values to construct a group initial direct - relation (IDR) matrix. However, it tends to be a low-precision expression, especially in complex practical problems. Although significant efforts have been made to improve the DEMATEL method, these improvements tend to neglect individual characteristics and group consensus, resulting in unconvincing decision results. This study provides a Dezert–Smarandache theory-based group DEMATEL method with reaching consensus. In order to reasonably determine the group IDR matrix, basic belief assignment function is employed to extract expert assessments and the proportional conflict redistribution rule no.5 of DSmT is employed to make fusion to derive the temporary group IDR matrix. Moreover, the consensus measures at both expert level and pair-factors level are calculated to determine whether the acceptable consensus level has been reached or not. If the required consensus level is not reached, a feedback mechanism will be activated to help experts reach a consensus. A consensus group IDR matrix for the group DEMATEL can be obtained with the help of feedback mechanism, based on which an algorithm is summarized for the proposed method to identify major factors in a complex system. Finally, numerical comparison and discussion are introduced to verify the effectiveness and applicability of the proposed method and algorithm.
ISSN:0926-2644
1572-9907
DOI:10.1007/s10726-019-09641-6