Consensus manipulation in social network group decision making with value-based opinion evolution

Opinion evolution is a common phenomenon in social network group decision-making (SNGDM), where the key issue is modelling trust weights among individuals. In social and economic networks, the links among individuals represent trust relationships that can offer value. However, extant studies on trust weight distribution have failed to consider individual network values. Moreover, in some scenarios, the external stakeholder may want to manipulate the opinion evolution process to achieve the desired consensus outcome. Therefore, this study proposes valuebased opinion evolution and discusses consensus manipulation in SNGDM. First, we measured the total value of an individual in a network, which consisted of two parts: intrinsic value and network value. Notably, we used the constant elasticity of substitution from economics as an operator to aggregate intrinsic and network values. Based on this, we described value-based opinion evolution and established an optimisation-based consensus manipulation model for SNGDM with value-based opinion evolution (CM-VBOE). Furthermore, we conducted some sensitive analyses, which indicate that an individual's value in the network increases with the decay factor and decreases with the elasticity of substitution, and provide a comparative analysis, which expands on the importance of accounting for the decay factor and elasticity of substitution when modelling value-based opinion evolution. Finally, a hypothetical application for crowdfunding project selection is presented to illustrate the applicability of the proposed CM-VBOE.

Automated summary

Opinion evolution is a common phenomenon in social network group decision-making. This study proposes value-based opinion evolution and discusses consensus manipulation in SNGDM. It establishes an optimisation-based consensus manipulation model and provides important insights on considering network values and external manipulation. Evaluation: 6 points.