Consistency-Driven Parameterized Prioritization Decision-Making Method with Generalized Probabilistic Linguistic Preference Information

The probabilistic linguistic preference relation (PLPR) enables decision-makers to express pairwise preferences with multiple linguistic terms and associated probabilities, effectively capturing uncertainty in evaluations. However, incomplete PLPRs and logical inconsistencies frequently arise in practice, compromising the reliability of rankings. This study proposes a parameterized consistency-driven prioritization method to restore missing information and improve logical coherence. First, generalized PLPRs (GPLPRs) are formally defined, and a deviation-tolerant completion model is introduced, in which a tunable parameter balances average and maximum deviations when estimating missing preference information. Then, considering the relationship between priority weights and consistency, a consistency-measurement approach is developed that integrates the enhancement of ranking discrimination and the filtering of noisy inputs with the combination of optimistic and pessimistic perspectives to reflect decision-makers’ evaluation attitudes. Further, a consistency-deviation-guided minimum distortion optimization model is proposed, ensuring targeted consistency improvement with minimal alteration of the original preferences, thereby forming a consistency-driven parameterized prioritization method for decision-making. Finally, two illustrative examples demonstrate the practical applicability of the proposed method to both individual and group decision-making contexts. Comprehensive parameter sensitivity and comparative analyses further validate the superiority of the proposed approach over existing approaches.