Abstract Due to the development of network technology, large-scale group decision making (LSGDM) has become increasingly concerned. In this paper, a k-core decomposition-based opinion leaders identifying method and clustering-based consensus model are developed for LSGDM problems. Firstly, a clustering method based on similarity degree is provided for dividing decision makers (DMs) into several clusters. Then, sub-clusters are presented for social networks (SNs) construction process, which are consist of DMs with same alternative ranking information. Furthermore, a novel k-core decomposition-based opinion leaders identifying method is proposed for selecting opinion leaders of these SNs. Finally, the opinion leaders identified are applied to the following clustering-based consensus model in LSGDM. The weights of DMs are distributed appropriately and the group can efficiently reach a consensus based on the proposed social network analysis (SNA) methods and consensus reaching process (CRP). A case study on flood disaster management shows that the proposed methods are feasible for LSGDM problems.

中文翻译：

---

一种基于k核分解的意见领袖识别方法和基于聚类的大规模群体决策共识模型

摘要 随着网络技术的发展，大规模群决策(LSGDM)越来越受到关注。在本文中，针对 LSGDM 问题开发了一种基于 k 核分解的意见领袖识别方法和基于聚类的共识模型。首先，提供了一种基于相似度的聚类方法，用于将决策者（DM）划分为多个聚类。然后，为社交网络 (SN) 构建过程呈现子集群，这些子集群由具有相同替代排名信息的 DM 组成。此外，提出了一种新的基于 k 核分解的意见领袖识别方法，用于选择这些 SN 的意见领袖。最后，将确定的意见领袖应用于 LSGDM 中以下基于聚类的共识模型。DM 的权重被适当分配，并且该组可以基于提出的社交网络分析 (SNA) 方法和共识达成过程 (CRP) 有效地达成共识。洪水灾害管理的案例研究表明，所提出的方法对于 LSGDM 问题是可行的。