Building a consensus for the best-worst method in group decision-making with an optimal allocation of information granularity

In a multiple-criteria group decision-making process, it may be unrealistic to assume that all experts are specialized in all aspects of the entire problem and can reach full agreement. To obtain a suitable result with more flexibility and robustness in group decision-making, this study concerns the introduction of information granules to the best-worst method model, regarded as an essential design asset to reach a consensus in a group decision-making scenario. More specifically, each pairwise comparison is performed through information granules instead of single numbers, by using an optimal allocation of information granularity. Further, we develop an optimization model that considers both consistency and consensus in one problem. The particle swarm optimization algorithm is used as an optimization method to find the optimal particle of the granular decision model. Furthermore, a new convergent iterative algorithm is developed to obtain the desired decision matrix. Next, an experimental study is presented to demonstrate the effectiveness, flexibility, and essence of the proposed model. Finally, a case study of autohome.com is conducted to select suitable automobiles for a company based on online reviews. The solving procedures are demonstrated, to further illustrate the feasibility of the proposed method.