Adaptive semi-supervised consensus model for multi-criteria large group decision making in a linguistic setting

In this paper we investigate the consensus reaching problem for Large Group Multi-Criteria Decision Making (MCLGDM). We present an adaptive, semi-supervised consensus model for MCLGDM problems with preferences expressed as Comparative Linguistic Expressions. Specifically, our work introduces an adaptive, semi-supervised feedback mechanism that, depending on the positions of decision makers' preferences and their level of uncertainty caused by hesitancy, requests human supervision to modify their preferences or updates them automatically. The proposed consensus model effectively handles large amounts of linguistic-natured information in consensus processes involving large groups. The methodology is illustrated and experimentally validated through a MCLGDM problem for candidate assessment in recruiting processes. Like-wise, a theoretical comparison with similar works is provided.