查看文章

Consensus reaching processes (CRPs) in group decision-making (GDM) problems seek a high level of collective agreement before making a decision. The presence of large groups of experts in such problems is increasingly frequent; therefore, it is necessary to propose new consensus models and approaches that reduce the high cost of the discussion process in these cases and to develop consensus support systems based on scalable architectures that facilitate the management of large groups computationally. This contribution presents a consensus support system for large-scale GDM problems, based on a multi-agent architecture that incorporates an agent-based semi-supervised autonomy approach aimed to minimize the cost invested in the CRP.