**Attraction Theory (AT)** establishes a computational framework for analyzing narrative engagement, applying the general cognitive psychology principles of Four-Domain Emergence and Dual-Track Transduction to the domain of storytelling. This theoretical work introduces four core innovations: (1) the **16-State Demand-State Network (DCN)** spanning physiological, psychological, and social domains as the foundation for narrative tension analysis; (2) **Dual-Track Decay functions** (φ and η) differentiating cognitive freshness decay (second-scale) from physiological emotional residue (minute-scale), explaining the independence of attention switching and emotional continuity; (3) the **ρ-Trajectory System** tracking audience immersion levels (ρ ∈ 0,1) across narrative phases with diagnostic thresholds for engagement fluctuations; and (4) **State-Competition Dynamics** modeling how competing demand-states vie for dominance via softmax mechanisms to determine the "first-state" governing audience attention at each narrative moment. By transforming narrative analysis from qualitative interpretation to quantitative dynamical systems analysis, AT provides a pathway for computational narrative analysis and generative storytelling. **Keywords**: Narrative Engagement, Demand-State Network, Dual-Track Decay, Substitution Coefficient, Computational Storytelling, Suspense Mechanics --- 吸引力理论(AT)建立了一个分析叙事吸引力的计算框架,将《四域涌现与双轨转导》的一般性认知心理学原理应用于故事创作领域。 本理论工作引入四项核心创新:(1)16态需求态网络(DCN),跨越生理、心理、社会三域的拓扑结构,作为叙事张力分析的基础;(2)双轨衰减函数(φ和η),区分认知新鲜度衰减(秒级)与生理情感残留(分钟级),解释注意力切换与情感连续性的独立性;(3)ρ-轨迹系统,追踪观众沉浸水平(ρ ∈ 0,1)随叙事阶段的演化,提供识别投入度波动的诊断工具;(4)态竞争动力学,建模竞争需求态如何通过softmax机制争夺主导权,决定每个叙事时刻的"首态"。 通过将叙事分析从定性解释转变为定量动力学系统分析,AT为计算叙事分析与生成式故事创作提供路径。 关键词:叙事吸引力,需求态网络,双轨衰减,代入系数,计算叙事,悬念机制
Bingqing Xie (Sun,) studied this question.