Two-Stage Feedback Mechanism With Different Power Structures for Consensus in Large-Scale Group Decision Making

This article investigates a two-stage consensus feedback mechanism that considers different power structures in large-scale group decision-making (LSGDM) environments. A Louvain algorithm type is used to detect subgroups in LSGDM by trust relationship. The concepts of internal subgroups and external subgroup consensus levels are defined, and an approach to identify the inconsistent individual/subgroup is developed to avoid the issue of “pseudoconsensus.” Combined with the background of the company’s shareholder power management regulations, the following three power structures are constructed: “absolute power,” “relative power,” and “democratic power.” A two-stage feedback mechanism is investigated with minimum adjustments to achieve the optimal power allocation under each power structure. This mechanism supports individuals reach consensus both inside and outside their subgroup. An illustrative example and discussions to verify the validity of the proposed method are reported.