Scholarlink

<p>Digital tourism marketing increasingly relies on professional user-generated content (PUGC), yet mechanisms through which multimodal videos and real-time social interactions shape travel decisions remain largely opaque. This study identifies these mechanisms by analyzing 2,650 tourism videos from Bilibili and their synchronized Danmaku (bullet-screen comments) through a computational framework integrating computer vision, natural language processing, and machine learning. We develop and validate five cognitive load metrics through comprehensive construct validity assessment (structural, known-groups, criterion, and nomological validity). Three findings emerge: First, the cognitive interaction term (SD × CD) ranks as the single most important predictor among 95 multimodal features (SHAP importance = 16.2%), with cognitive features collectively accounting for 44.5% of top-20 predictive importance, showing that cognitive engagement is the central mediating mechanism through which multimodal content generates travel intention. Second, non-linear models substantially outperform linear baselines (<i>R</i>² = 0.436 vs. 0.033), with construct-validated cognitive load metrics alone achieving test <i>R</i>² = 0.355 and the extended cognitive model (incorporating interaction terms) achieving bootstrap <i>R</i>² = 0.343 (95% CI: [0.286, 0.395]), revealing threshold effects and cognitive gating mechanisms not captured by additive frameworks. Third, Danmaku functions as cognitive scaffolding rather than distraction, with Granger causality suggesting orchestrated attention cascades from visual to audio to textual processing (<i>p</i> &lt; 0.001). By operationalizing previously unobservable mechanisms through extraction of 95 multimodal features aligned with 4.75&#xa0;million Danmaku comments, this study advances understanding of platform affordances and tourism decision-making within a Computationally Extended S-O-<i>R</i> framework.</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

From watching to wishing: a multimodal computational analysis of how PUGC video-Danmaku ecology shapes travel intention

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月26日

查看全文
  • Feng Ye,
  • Min Yin,
  • Shouqian Sun,
  • Xuanzheng Wang

摘要

Digital tourism marketing increasingly relies on professional user-generated content (PUGC), yet mechanisms through which multimodal videos and real-time social interactions shape travel decisions remain largely opaque. This study identifies these mechanisms by analyzing 2,650 tourism videos from Bilibili and their synchronized Danmaku (bullet-screen comments) through a computational framework integrating computer vision, natural language processing, and machine learning. We develop and validate five cognitive load metrics through comprehensive construct validity assessment (structural, known-groups, criterion, and nomological validity). Three findings emerge: First, the cognitive interaction term (SD × CD) ranks as the single most important predictor among 95 multimodal features (SHAP importance = 16.2%), with cognitive features collectively accounting for 44.5% of top-20 predictive importance, showing that cognitive engagement is the central mediating mechanism through which multimodal content generates travel intention. Second, non-linear models substantially outperform linear baselines (R² = 0.436 vs. 0.033), with construct-validated cognitive load metrics alone achieving test R² = 0.355 and the extended cognitive model (incorporating interaction terms) achieving bootstrap R² = 0.343 (95% CI: [0.286, 0.395]), revealing threshold effects and cognitive gating mechanisms not captured by additive frameworks. Third, Danmaku functions as cognitive scaffolding rather than distraction, with Granger causality suggesting orchestrated attention cascades from visual to audio to textual processing (p < 0.001). By operationalizing previously unobservable mechanisms through extraction of 95 multimodal features aligned with 4.75 million Danmaku comments, this study advances understanding of platform affordances and tourism decision-making within a Computationally Extended S-O-R framework.