Towards a Fairness-Oriented Approach to Consensus Reaching Support Under Fuzzy Preferences and a Fuzzy Majority via Linguistic Summaries

A novel approach to a human centric support of a consensus reaching process in a group of agents who present their testimonies as individual fuzzy preference relations is proposed. The concept of a degree of consensus is used which is meant as the degree to which, for instance, most of important agents agree as to almost all of relevant options. The fuzzy majorities are equated with linguistic quantifiers and Zadeh's calculus of linguistically quantified propositions is used. The new concepts of a consensory and dissensory agent is introduced. The authors' approach of using linguistic data summaries for a comprehensive summarization of how the agents' current testimonies look like is then employed for the consensory and dissensory agents to obtain suggestions to the agents on changes of specific preferences that could lead to a higher degree of consensus. An explicit inclusion of opinions of the consensory and dissensory agents is shown to be an important step towards a fairness type attitude of the moderator as opinions of all agents are accounted for.