Group Decision-Making with Linguistic Cognition from a Reliability Perspective

To deal reliably with the cognitive uncertainty experienced by decision-makers when facing problems involving linguistic group decision-making, we investigate a new research perspective: cognitive familiarity is regarded as a measure of cognitive reliability. The linguistic variables examined in this work are quantified with the use of several granulation optimization models that include consideration of cognitive reliability. Three types of linguistic variables are used for describing the alternative grades, attribute weights, and levels of cognitive familiarity associated with the experts involved. Three interrelated optimization models are built to quantify these linguistic variables. An information entropy model first determines cognitive familiarity, which is applied as a measure of cognitive reliability. Using group consistency and cognitive reliability, two proposed optimization models then successively establish the attribute weights and alternative grades. The final element is a proposed new selection method based on the aggregated values for the alternative grades and cognitive reliability. An illustrative example clarifies the steps in the proposed method, which produces a ranking of three alternatives as a final decision. The validity and advantages of the proposed method are verified through a comparison with existing approaches. The proposed method can be employed for effectively resolving decision-making uncertainty through the improvement of the group consistency and cognitive reliability of the experts. Sensitivity analysis also reveals that cognitive reliability has a strong impact on decision-making and should thus be considered during fusion processes.