Social Network Group Decision Making Considering Order Effects with Quantum Cognition

Social network group decision making generally involves experts who have complex social relationships to influence each other. Order effect arises from the order of opinion representation among decision makers. Nowadays, quantum theory has become an efficient instrument for simulating and quantifying the complex behavior among the decision makers. This paper proposes a quantum framework for modelling order effect in social network group decision making process. Firstly, we integrate the opinion similarity into social connections. Then, the Louvain Algorithm is used to cluster experts into different subsets. The weights of individuals and clusters are obtained by using a constructed sociability-similarity centrality index. And we fuse individual opinions into a collective one within each cluster. Next, the path probability method in quantum framework is used to aggregate opinions of all clusters, which interprets order effect by converting a state vector with different sequences of operators to represent different orders of opinion representation. Then, we introduce an algorithm to determine a local optimal order of opinion representation to minimize order effect among clusters. In the end, we apply the proposed model to a numerical example and demonstrate its effectiveness and stability through detailed sensitivity analysis and comparative analysis.