A Feedback Mechanism With Unknown Bounded Confidence-Based Optimization Model for Consensus Reaching in Social Network Group Decision Making
针对群体决策中决策者过度自信的问题，提出基于客观有界置信度的反馈机制，利用分布语言偏好关系构建共识模型，并通过优化方法提供建议，提升共识效率。
Various feedback mechanisms focus on bounded confidence in the consensus reaching process (CRP) for group decision making (GDM) problems. However, confidence level from DMs’ subjective cognition can lead to over-confidence, and thus to have negative effect on CRP. With this idea in mind, this article proposes an objective way to determine bounded confidence levels. In this article, the distribution linguistic preference relation (DLPR) is used to describe decision makers’ (DMs’) preferences on alternatives. A consensus reaching model with DLPRs in social network GDM (SNGDM) with bounded confidence effect is constructed. In the proposed consensus approach, the objective bounded confidence level is obtained from individual professional performance and social performance, i.e., knowledge degree based on consistency index and entropy measure of DLPRs, and the reliability degree based on trust degree received from other DMs. Then, the acceptable advices based on a bounded confidence-based optimization approach is provided for the identified DMs. Finally, a numerical example and comparative simulation analysis are provided to justify its feasibility and superiority.