Personalized individual semantics derived consensus model in hesitant fuzzy linguistic MCGDM based on discrimination degrees and multidimensional preferences

In multi-criteria group decision making (MCGDM) with qualitative settings, hesitant fuzzy linguistic term sets (HFLTSs) provide a flexible way to capture decision makers (DMs)’ hesitancy when eliciting linguistic expressions. However, existing studies overlook the fact that words mean different things to different people, which entails that DMs have personalized individual semantics (PISs) in terms of their expressions in linguistic MCGDM. This study develops a novel framework to address hesitant fuzzy linguistic MCGDM considering PISs of DMs. First, the concept of discrimination measure for DMs is defined. Based on the discrimination measure, a discrimination-based optimization model and a multidimensional preference-based optimization model are established to derive personalized numerical scales (NSs) of linguistic terms for DMs in different situations. Second, a consensus-reaching method based on an optimization model that aims to minimize the amount of adjustments between the original and updated linguistic decision matrices and to preserve their accuracy is constructed to yield a consensual solution. Finally, an illustrative example followed by comparative and sensitivity analysis is presented to demonstrate the application and features of the proposed framework in this study.