Premium
An overlap graph model for large‐scale group decision making with social trust information considering the multiple roles of experts
Author(s) -
Liao Huchang,
Tan Runzhi,
Tang Ming
Publication year - 2021
Publication title -
expert systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.365
H-Index - 38
eISSN - 1468-0394
pISSN - 0266-4720
DOI - 10.1111/exsy.12659
Subject(s) - computer science , graph , group decision making , rationality , construct (python library) , decision analysis , process (computing) , decision maker , scale (ratio) , operations research , psychology , theoretical computer science , social psychology , mathematics , quantum mechanics , political science , law , programming language , operating system , statistics , physics
Social network analysis is an efficient tool to investigate the relationships of decision‐makers in large‐scale group decision making (LSGDM). Existing social network‐based LSGDM studies generally assumed that each decision‐maker has a single role or belongs to only one subgroup. The assumption that a decision‐maker has multiple roles or belongs to multiple subgroups is rarely taken into consideration. In this regard, this study proposes an overlap graph model (OGM) in which decision‐makers can participate in the decision‐making process in multiple roles to solve LSGDM problems with social trust information. In the OGM, decision‐makers are firstly divided into two types: multiple‐role decision‐makers and single‐role decision‐makers. Since it is unpractical for a decision‐maker to evaluate all others in a LSGDM problem, we then investigate how to construct a complete social trust network based on an Agent mechanism. A two‐stage consensus reaching process is proposed to reduce the discrepancies among decision‐makers: The first stage is for single‐role decision‐makers within a subgroup while the second stage is for Agents and multiple‐role decision‐makers. Finally, an illustrative example regarding selecting treatment plans for critical patients in COVID‐19 is provided to test the applicability and rationality of the proposed model.