Multi-criteria group sorting (MCGS) concerns the multi-person process of evaluating alternatives and assigning them to predetermined ordered categories under certain criteria. Identifying class thresholds and coordinating conflicting opinions are two primary challenges within this process. Taking into account the 2-tuple linguistic preferences of decision-makers (DMs), we propose a consensus-oriented sorting method that incorporates dynamic trust management for linguistic MCGS. The main contributions and innovations are as follows. (a) A group consensus maximization model is constructed to optimize class thresholds and initial individual sorting results. (b) Integrating familiarity-based trust and similarity-based trust, the dynamic comprehensive trust among DMs is established. On this basis, a feedback mechanism is designed, in which the bi-objective programming is utilized to generate adjustment recommendations. (c) The final consensual group sorting result is obtained by maximizing the similarity of group views. (d) The feasibility of the proposed method is demonstrated using a case of sorting fresh agricultural products. Using sensitivity analysis and comparative analysis, the proposed method is verified to be rational and effective. A comparison based on real datasets and a qualitative comparison further demonstrate the validity and characteristics of the proposed method.

中文翻译：

---

结合动态信任管理的面向共识的语言多标准分组排序方法


多标准组排序 (MCGS) 涉及评估备选方案并根据特定标准将其分配到预定有序类别的多人过程。确定类别阈值和协调相互冲突的意见是此过程中的两个主要挑战。考虑到决策者（DM）的二元组语言偏好，我们提出了一种面向共识的排序方法，该方法结合了语言 MCGS 的动态信任管理。主要贡献和创新如下。 (a) 构建群体共识最大化模型来优化类别阈值和初始个体排序结果。 (b)融合基于熟悉度的信任和基于相似性的信任，建立了DM之间的动态综合信任。在此基础上，设计了反馈机制，利用双目标规划生成调整建议。 (c)通过最大化群体观点的相似度来获得最终一致的群体排序结果。 (d)通过生鲜农产品分选案例证明了该方法的可行性。通过敏感性分析和比较分析，验证了所提方法的合理性和有效性。基于真实数据集的比较和定性比较进一步证明了该方法的有效性和特点。