Scholarlink

<p>Power asymmetry is a common feature of conflicts involving multiple decision makers (DMs), where differences in influence, objectives, and behavioral tendencies often lead to strategic imbalance and unstable outcomes. The Graph Model for Conflict Resolution (GMCR) provides an effective framework for analyzing such interactions. However, existing studies addressing power asymmetry in GMCR usually assign the relative power of DMs subjectively, lacking an objective basis for power structure division. To overcome this limitation, this paper incorporates community division into the GMCR framework and proposes an improved cosine-similarity spectral clustering algorithm. Since DMs frequently express their preferences using linguistic terms, and the strength of these expressions can differ across power levels, traditional GMCR models are not well suited to handle such information. Linguistic distribution assessment (LDA) is therefore used to extract probabilistic linguistic preferences while retaining their descriptive characteristics. However, traditional ranking methods have difficulty reflecting the power differences among DMs. Therefore, superiority and inferiority ranking (SIR) is integrated with LDA to form an LS (LDA–SIR, abbreviated as LS) environment, enabling multi-attribute strategy ranking that reflects power differences and provides unified preference representation for composite DMs (CDMs). In addition, four new stability definitions are developed under the LS environment to enhance equilibrium analysis and strengthen theoretical interpretability. Finally, the proposed GMCR-based framework provides a general tool for analyzing multi-party strategic conflicts, illustrated through a representative case of inter-enterprise competition in the policy-driven new energy automobile industry.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Community Structure-Based GMCR Framework for Power-Asymmetric Conflict Resolution Under Linguistic Distribution Information

  • 期刊论文

  • 发表日期：

    2026年4月3日

查看全文
  • Peide Liu,
  • Xueke Wang,
  • Qian Pan,
  • Peng Wang

摘要

Power asymmetry is a common feature of conflicts involving multiple decision makers (DMs), where differences in influence, objectives, and behavioral tendencies often lead to strategic imbalance and unstable outcomes. The Graph Model for Conflict Resolution (GMCR) provides an effective framework for analyzing such interactions. However, existing studies addressing power asymmetry in GMCR usually assign the relative power of DMs subjectively, lacking an objective basis for power structure division. To overcome this limitation, this paper incorporates community division into the GMCR framework and proposes an improved cosine-similarity spectral clustering algorithm. Since DMs frequently express their preferences using linguistic terms, and the strength of these expressions can differ across power levels, traditional GMCR models are not well suited to handle such information. Linguistic distribution assessment (LDA) is therefore used to extract probabilistic linguistic preferences while retaining their descriptive characteristics. However, traditional ranking methods have difficulty reflecting the power differences among DMs. Therefore, superiority and inferiority ranking (SIR) is integrated with LDA to form an LS (LDA–SIR, abbreviated as LS) environment, enabling multi-attribute strategy ranking that reflects power differences and provides unified preference representation for composite DMs (CDMs). In addition, four new stability definitions are developed under the LS environment to enhance equilibrium analysis and strengthen theoretical interpretability. Finally, the proposed GMCR-based framework provides a general tool for analyzing multi-party strategic conflicts, illustrated through a representative case of inter-enterprise competition in the policy-driven new energy automobile industry.