Mining Personalized Individual Semantics of Self-confidence Participants in Linguistic Group Decision-Making

Abstract

Words representing individual preferences in group decision-making (GDM) are always associated with different meanings. Consequently, mining personalized semantics of decision-makers (DMs) hidden in preference expressions, and establishing a corresponding management mechanism, is an effective way to reach group consensus through computing with word methodology. However, the aforementioned consensus-reaching process may be hindered by self-confidence. To address this limitation, this study proposes a linguistic group decision model with self-confidence behavior. First, we identified the corresponding self-confidence levels for each DM. Next, we integrated different linguistic representation models into unified linguistic distribution-based models. We then obtained individual personalized semantics based on a consistency-driven optimization method, and designed a feedback-adjustment mechanism to improve the adjustment willingness of DMs and group consensus level. Finally, we conducted a quantitative experiment to demonstrate our model’s effectiveness and feasibility.