Scholarlink

<p>This study proposes a multimodal biosensing and AI-driven dynamic tuning system to address the challenge of belief transformation in physical education (PE) among rural students. By integrating physiological (e.g., heart rate variability), behavioral (e.g., collaboration error rate), and cognitive (e.g., reflective journal semantics) data, the system employs a Responsibility Attribution Reinforcement (RAR) framework to operationalize collectivist cultural elements into measurable behavioral anchors. A Random Forest-based algorithm enables real-time adaptive adjustment of task difficulty and intervention intensity via edge computing nodes, ensuring low-latency responses (&lt; 500 ms) in resource-constrained environments. Experimental results from a quasi-experimental study (<i>N</i> = 98) demonstrate significant improvements in academic belief scores (d = 0.73, <i>p</i> &lt; 0.001) and attribution strategy optimization (78.3%, <i>p</i> &lt; 0.001), supporting the potential efficacy of the closed-loop “perception–analysis–response” mechanism. The system offers a scalable, culturally attuned solution in the studied context for enhancing belief internalization and educational equity in rural settings.</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 rural student sports belief transformation system based on multimodal biosensing and AI dynamic tuning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月22日

查看全文
  • Xinyu Fu,
  • Ze Meng,
  • Heng Shi

摘要

This study proposes a multimodal biosensing and AI-driven dynamic tuning system to address the challenge of belief transformation in physical education (PE) among rural students. By integrating physiological (e.g., heart rate variability), behavioral (e.g., collaboration error rate), and cognitive (e.g., reflective journal semantics) data, the system employs a Responsibility Attribution Reinforcement (RAR) framework to operationalize collectivist cultural elements into measurable behavioral anchors. A Random Forest-based algorithm enables real-time adaptive adjustment of task difficulty and intervention intensity via edge computing nodes, ensuring low-latency responses (< 500 ms) in resource-constrained environments. Experimental results from a quasi-experimental study (N = 98) demonstrate significant improvements in academic belief scores (d = 0.73, p < 0.001) and attribution strategy optimization (78.3%, p < 0.001), supporting the potential efficacy of the closed-loop “perception–analysis–response” mechanism. The system offers a scalable, culturally attuned solution in the studied context for enhancing belief internalization and educational equity in rural settings.