社会网络环境下基于负面行为管理与改进的最小成本共识模型的大群体决策方法

Large-scale group decision-making method based on negative behavior management and improved minimum cost consensus model in social network environment

提出基于社会网络分析的专家负面行为识别与管理,以及基于改进的最小成本共识模型的大群体决策方法.首先,定义专家正面社会形象、负面社会形象以及观点指标等概念,根据专家负面行为表现方式提出负面社会形象管理、负面评估偏好管理以及双重负面行为管理3种策略;其次,计算专家的观点相似度、正面社会形象相似度、负面社会形象相似度,基于K均值聚类方法对大群体专家聚类;然后,考虑社会形象对共识达成的影响作用,提出基于社会形象的改进最小成本共识模型,以达成群体共识并获得最终方案排序;最后,通过一个众筹平台选择算例分析说明所提出方法的有效性和可行性.

This paper proposes expert negative behavior identification and management based on social network analysis,and constructs an improved minimum cost consensus model in the large-scale group decision-making method.First,the concepts of experts'positive social image,negative social image and opinion indicator are defined,and three management strategies including negative social image,negative evaluation preference and double negative behavior to manage experts negative behavior are proposed.Secondly,the similarity of opinions,the similarity of positive social image,and the similarity of negative social image of experts are calculated,and large-scale experts are clustered based on the K-means clustering method.Then,considering the influence of social image on consensus,an improved minimum-cost consensus model based on social image is proposed to reach group consensus and obtain the final plan ranking.Finally,a crowdfunding platform selection is presented to show the effectiveness and feasibility of the proposed method.