Analysis and extension of ordinal priority approach for multi-attribute group decision-making problems

초록

The ordinal priority approach (OPA) model, which requires only ordinal preference rankings as input, is used to assess the weights of alternatives, attributes, and experts in multi-attribute group decision-making problems. In this paper, we revisit the OPA model and present a closed-form solution derived through convex analysis. From a methodological perspective, this analysis not only enhances computational convenience by eliminating the need for linear programming software, but also provides deeper insight into the underlying structure of the OPA model. In addition, we introduce a decreasing convex sequence of weights among ranked alternatives. From a model refinement standpoint, this further restricts the feasible region of the original OPA model, thereby facilitating a clearer identification of preferred alternatives. Finally, from the perspective of attribute structure, we propose a new OPA model with a hierarchical structure of attributes, which broadens the model's applicability to a wider range of real-world decision-making problems, as demonstrated through a supplier selection case study.

키워드