A two-stage ranking method to minimize ordinal violation for pairwise comparisons

摘要: Abstract Pairwise comparison is a powerful tool in intelligent decision making systems. Items are compared using numerical judgments that estimate the item weight ratios, which provided by makers or transformed objective data. The reliable assignment of important, because judgment variety leads to significantly different results. However, it difficult provide exact ratios for owing limitations knowledge. Although data provides an estimation series artificially defined rules required transform into judgments; however, these subjective and arbitrary. Conversely, dominance relationships between items obvious reliable. Therefore, this study proposes two-stage ranking method minimize ordinal violation indicates degree conflict result dominance. First, 0–1 integer programming designed solved. Then, second stage focuses on topological sorting nodes graph constructed optimal solution. To validate effectiveness proposed method, we perform two experiments: example participants real-world application involving top tennis players. results show not only avoids subjectivity judgments, but also obtains has minimum among methods.