An approach based on computing with words to manage experts behavior in consensus reaching processes with large groups

Group decision making problems are characterized by the participation of multiple experts with different points of view, who attempt to find a common solution to a problem composed by a set of alternatives. Such problems are often defined in environments of uncertainty caused by the imprecision and vagueness of information, therefore experts must utilize appropriate information domains to deal with such uncertainty when expressing their preferences, e.g. linguistic information. Usually, in group decision making problems it is necessary to apply a consensus reaching process, in which experts discuss and make their opinions closer to each other, in order to achieve a high level of agreement before making the decision. Nevertheless, in large-scale group decision making problems, where a large group of individuals take part, it is more frequent the existence of certain subgroups with a non-cooperative behavior towards consensus reaching. For this reason, it would be convenient to identify such subgroups and deal with them, so that their behavior does not affect the consensus reaching process negatively. In this contribution, we present an approach based on computing with words and fuzzy set theory, to study the behavior of experts in consensus reaching processes, with the aim of identifying and penalizing the importance weights of those experts whose behavior does not contribute to reach a collective agreement.