A heterogeneous multi-attribute case retrieval method for generating emergency response alternatives based on interval evidential reasoning and gained and lost dominance score method: An application to the COVID-19 emergency situation

Abstract

Heterogeneity exists in the case information of an emergency, and case retrieval plays a crucial role in generating an emergency response alternative. Despite advancements, fusing heterogeneous case information and considering decision maker preferences remain challenging issues to be solved. To address this, a novel heterogeneous multi-attribute case retrieval method is proposed and applied to COVID-19 emergency decision-making in this study. To overcome uncertainties and heterogeneities in case information, a new interval evidential reasoning (IER)-based case similarity measurement is developed to better protect data intergrity and improve case retrieval accuracy. Next, the best-worst method and information entropy are used to obtain subjective and objective weights of evaluated attributes, respectively, and subsequently these weights are integrated. Furthermore, a novel interval gained and lost dominance score (GLDS) method based on TODIM (Portuguese acronym for interactive multi-criteria decision-making) is developed to derive the comprehensive utilities of the similar historical case. This method simultaneously considers the fairness of the decision-making process and the psychological behavior of decision makers. Finally, a simulation example with large-scale data and two illustrated examples of generating emergency response altenatives are provided to demonstrate the applicability of the proposed method. Sensitivity and comparative analysis are conducted to verify the validity of the method.