Researches And Applications Of Group Decision Making Based On Opinion Dynamics

Decision making activities widely exist in enterprise management.With the rapid development of mobile internet,decision-making activities become more and more complex.Individual decision-making can not meet the daily decision-making activities of enterprises.Therefore,multiple decision-makers need to participate in group decision-making.The emergence of social media enables people to communicate on the network,and build the relationships between people through the network carrier,so as to form a social network with certain characteristics.Therefore,analyzing the group opinion dynamics pattern in social networks and constructing group information evaluation model is of great significance to enterprise management.Due to the different backgrounds of decision makers,the preference forms given are also diverse.It is very important to select appropriate and efficient aggregation operators to make decisions.When the opinions of decision-makers differ greatly,direct information aggregation will lead to the irrationality of decision-making results.In addition,the group decision-making process in the social network is a complex dynamic process.The existing researches ignore the impact of the social network relationships between decision-makers on the opinions of decision-makers.Based on this,this paper focuses on the theory and method of group decision-making in enterprise management.Based on the preference information of decision-makers,this paper explores the opinion dynamics behaviors of decision-makers under social networks,and analyzes the impact of social networks on consensus achievement efficiency and decision results in group decision-making.The main contents of this paper are as follows:(1)Heterogeneous group preference information aggregation method based on matrix similarity.Firstly,different heterogeneous preference information is transformed into fuzzy preference relation matrix by different transformation functions.Then,an improved power weighted average operator based on matrix similarity is proposed,and some properties of the operator are given.Finally,based on the proposed improved power weighted average operator,a new preference information aggregation method for heterogeneous groups is constructed,and the ranking method based on complementary judgment matrix is used to select the best decision scheme.Finally,a numerical example is given to verify the effectiveness and feasibility of the method proposed in this paper.(2)Multi-attribute group decision making based on De Groot model.In order to analyze the impact of social relationships between decision makers on decision results,a multi-attribute group decision making method based on social trust network is proposed.The method includes three stages: trust propagation,consensus reaching process and scheme selection.In the trust propagation stage,a complete social trust matrix is constructed based on trust propagation and trust aggregation,and the decision-maker's self-confidence value is introduced to obtain the social weight influence matrix and the decision-maker's weight.In the process of reaching consensus,the consensus measure is used to determine whether there is consensus among decision makers,and the feedback mechanism based on De Groot model is used to adjust the opinions of decision makers who do not reach consensus.When the decision-makers reach a consensus,they get the group opinions according to the aggregation operator and sort it to select the best scheme.Finally,the effectiveness of this method is verified by the numerical experiment.The comparative analysis shows that the proposed method can improve the decision efficiency compared with the group decision-making method without considering trust propagation and trust aggregation.(3)Large scale group decision making based on bounded confidence and social networks.The diversity and complexity of information make more decision-makers participate in the decision making activities and the scale of decision-making is becoming larger and larger.How to use the differences of opinions and trust relationships among decision makers to promote large-scale group consensus is an important problem.This paper proposes a large-scale group decision-making method based on bounded confidence and social network to management decisionmakers' opinions.Firstly,the fast unfolding algorithm is used to reduce the dimension of large-scale groups,and the weight of decision-makers is obtained through social network analysis.Secondly,a consensus reaching process model based on Manhattan distance is established and a feedback mechanism based on social network and bounded confidence is constructed to help decision makers reach consensus.Finally,the feasibility of this method is verified by a numerical example.The comparative analysis results show that the proposed method can make the speed of reaching consensus faster and the efficiency of information management higher.(4)A supplier selection group decision-making method based on multi-stage fuzzy information and decision-makers' opinion interaction.Since the opinions of decision makers usually fluctuate over time,decision makers need to evaluate alternatives in the past several periods rather than a single period.With the change of environment,decision makers will exchange opinions and influence each other.In order to meet the different evaluation requirements of decision makers and analyze the influence of time factors and opinion interaction among decision makers,a group decision-making method considering multi-stage fuzzy information and opinion interaction of decision makers is proposed.In this method,decision makers use generalized fuzzy numbers to provide their preferences in multiple stages,and determine the weights of different stages through mathematical programming method.The influence of opinion interaction among decision makers is considered by assigning different weights to different decision makers.Finally,the ideal solution similarity method is used to rank the potential suppliers.The results of supplier selection example show that this method can consider multi-stage fuzzy information and opinion interaction to select the appropriate supplier.Group decision-making based on the decision-makers' preference information and opinion dynamics is a practical problem faced by enterprise management in the era of social networks and big data.Through in-depth research on this problem,it will help to expand the traditional group decision-making methods to the social network environment and further improve the existing method system of group decision-making in enterprise management.