An Enhanced Decision-Making Framework Driven by Complex Semantics under Nested Probabilistic Linguistic Environments

Weidong Gan,Zeshui Xu,Xinxin Wang

DOI: https://doi.org/10.1007/s12559-024-10245-9

IF: 4.89

2024-01-01

Cognitive Computation

Abstract:Language is the most important way of information transmission for human beings, through which people exchange ideas and guide practice. But the semantic expression has uncertainty, which usually brings difficulties to human’s decision-making behaviors in routine life. Hence, exploring methods to effectively handle the uncertain cognitive information has been an enduring research topic in recent years. Nested probabilistic linguistic term set (NPLTS) and ELECTRE II, as a powerful linguistic model and a classical decision-making method respectively, provide us useful tools. But the extant studies have some noticeable defects. Thus, this work proposes an enhanced decision-making framework in nested probabilistic linguistic environments. We refine the comparison method of NPLTSs by using a novel calculation method of the performance score and confusion degree. And we propose a new distance measurement method based on a symmetric linguistic scale. Considering the objective and subjective weights of criteria, we establish an entropy-based model to calculate the comprehensive weights of criteria. Then, we refine the classical ELECTRE II method. Based on which, we develop an enhanced ELECTRE II framework. The proposed model successfully addresses the potential failure of the classical ELECTRE II method. Based on comparisons with other methods, the alternative ranking changes in response to the fluctuation of criteria weights while maintaining rationality. Thus, it is proved that the proposed model is of excellent effectiveness, good reliability, and appropriate sensitivity.