Multi-Granular Large Scale Group Decision-Making Method with a New Consensus Measure Based on Clustering of Alternatives in Modifiable Scenarios.

Because a decision-making process in the real world involves more people and it is more heterogeneous nowadays, this paper develops an innovative large scale multi-granular decision-making method that is intended to be used in situations where many experts are involved, who do not have to provide their opinions about the alternatives at the same time. The proposed method has three main characteristics. Firstly, experts provide their judgments using the numerical labels set of their choice. Secondly, experts can join the decision-making process in any of its rounds. Thirdly, this method performs a clustering that sorts the experts according to their judgments, obtaining how many experts belong to the same group and the consensus among them. These two parameters allow the weight of the group to be adjusted. Finally, the method uses inter-group consensus measures to verify that the groups agree with the decision made.