Group Decision Making Based on a Framework of Granular Computing for Multi-Criteria and Linguistic Contexts

The usage of linguistic information involves computing with words, a methodology assuming linguistic values as computational elements, in group decision-making environments. In recent times, a new methodology founded on a framework of granular computing has been employed to manage linguistic information. An advantage of this methodology is that the distribution and the semantics of the linguistic values, in place of being initially established, are defined by the optimization of a certain criterion. In this paper, different from the existing approaches, we present a novel approach build on the basis of a granular computing framework that is able to cope with group decision-making problems defined in multi-criteria contexts, that is, those in which different criteria are considered to evaluate the possible alternatives for solving the problem. In particular, it models group decision-making problems in a more realistic way by taking into account that each criterion has an importance weight and by considering that each decision maker has a different importance weight for each criterion. This approach makes operational the linguistic values by associating them with intervals via the optimization of an optimization criterion composed of two important aspects that must be taken into account in this kind of decision problems, that is, the consensus at the level of group of decision makers and the consistency at the level of individual decision makers.