Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

The DS/AHP Method of MCDM under Ignorance: Vagueness, Consideration Sets and Consensus Building in a Group Decision Problem

Beynon, Malcolm James ORCID: https://orcid.org/0000-0002-5757-270X 2004. The DS/AHP Method of MCDM under Ignorance: Vagueness, Consideration Sets and Consensus Building in a Group Decision Problem. Presented at: The 17th International Conference on Multiple Criteria Decision Making, Whistler, BC, Canada, 6-11 August 2004.

Full text not available from this repository.

Abstract

DS/AHP is a nascent method of multiple criteria decision making (MCDM) that operates in the presence of ignorance. The notion of ignorance in this case may describe hesitancy and doubt in making decisions due to a ‘consequential action’ nature of the problem considered. The mathematical foundation of DS/AHP is based on Dempster-Shafer theory (DST), considered a generalisation of the more traditional Bayesian theory. It adopts a hierarchical structure to deconstruct a problem similar to the Analytic Hierarchy Process. This paper further exposits the use of DS/AHP in the area of group MCDM, through a hypothetical example. Each group member (DM) is able to make preference judgements on groups of decision alternatives (DAs), as well as not judging on all the DAs. For a single criterion the judgements by a DM are expressed as a body of evidence - BOE (from DST). A BOE contains levels of exact belief (mass) values on the preferment of groups of DAs and a level of concomitant ignorance. The structure of a BOE allows a variety of information describing a DM and the group in their decision making. Decision making intensity at the individual DM level is first considered. This relates to the levels of local ignorance and non-specificity in the judgements made by a DM on each criterion. Here, associated unconstrained and constrained index values are evaluated, which quantify local ignorance and non-specificity and plotted in a recently defined vagueness plot. The results allow a concise visual interpretation to the intensity in the inherent judgement making of DMs and their clustering. The next issue, again at the individual DM level relates to the identification of the most preferred DAs. Two DST related measures, belief and plausibility, offer alternative approaches to identify groups of DAs of different sizes most preferred. Within consumer behavior the belief and plausibility measures relate to identifying those groups which are most supported (belief) and the compliment of those least supported (plausibility). A number of graphs are displayed which elucidate the identified groups of DAs of different sizes. A direct analogy is shown with the choices of a consumer and their awareness, consideration and choice sets, is discussed. The final issue investigated, considers the construction of a final group decision. The order of combination of the BOE from group members using a rule based on a DST related distance measure is described. The memory retaining property of BOE combination allows the construction process to be considered a consensus building approach. The whole approach is elucidated through a dendogram and further identifies the possibility of a majority rule decision. Throughout the study the opportunities for the visual representation of findings allows a clear understanding to the processes inherent in group decision making.

Item Type: Conference or Workshop Item (Paper)
Date Type: Completion
Status: Unpublished
Schools: Business (Including Economics)
Subjects: H Social Sciences > H Social Sciences (General)
H Social Sciences > HD Industries. Land use. Labor > HD28 Management. Industrial Management
Q Science > QA Mathematics
Uncontrolled Keywords: Dempster-Shafer Theory; Consensus building; Ignorance; Group decision making; MCDM
Related URLs:
Last Modified: 21 Oct 2022 09:47
URI: https://orca.cardiff.ac.uk/id/eprint/37682

Actions (repository staff only)

Edit Item Edit Item