Scholarlink

<p>This study examines gender- and age-specific dynamics of non-work travel in Guangzhou, China, using cellphone data from 6&#xa0;million users combined with interpretable machine learning. Although non-work travel constitutes a major share of daily mobility, its life-course and caregiving dimensions remain underexplored. A Gradient Boosting Decision Tree (GBDT) model with SHapley Additive exPlanations (SHAP) is applied to analyze individual non-work travel distances and interpret associations with gender, age, and destination facilities. Several contributions emerge. Conceptually, we uncover life-stage specific gender mobility patterns that were previously obscured, including clear threshold effects in age. In particular, we identify critical age breakpoints (around 28 and 48 years) where the relationship between age and travel distance shifts direction. This reveals a “gender-masking” effect: aggregate analyses would suggest a negligible gender influence, whereas our approach shows that young women travel farther than young men, but active-caregiving women travel far less than men due to caregiving constraints. Besides, we introduce the notion of a “gender difference amplifier”, quantifying how certain destinations disproportionately extend one gender’s travel over the other. For instance, shopping facilities significantly amplify women’s travel distance during caregiving years, while cultural dining traditions pull post-intensive-caregiving men’s travel farther. These nuanced insights—such as caregiving obligations curbing women’s mobility and social traction drawing post-intensive-caregiving men – were hard to discern with earlier data or methods. By combining big data with interpretable machine learning, this study provides new theoretical and empirical insights into how gendered life-course roles and urban opportunity structures interact to shape everyday non-work mobility.</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

Intersecting gender and age in non-work travel: evidence from cellphone data and interpretable machine learning in Guangzhou, China

  • 期刊论文

  • 发表日期：

    2026年3月3日

查看全文
  • Zifeng Chen,
  • Xinyi Ding,
  • Ying Zhao

摘要

This study examines gender- and age-specific dynamics of non-work travel in Guangzhou, China, using cellphone data from 6 million users combined with interpretable machine learning. Although non-work travel constitutes a major share of daily mobility, its life-course and caregiving dimensions remain underexplored. A Gradient Boosting Decision Tree (GBDT) model with SHapley Additive exPlanations (SHAP) is applied to analyze individual non-work travel distances and interpret associations with gender, age, and destination facilities. Several contributions emerge. Conceptually, we uncover life-stage specific gender mobility patterns that were previously obscured, including clear threshold effects in age. In particular, we identify critical age breakpoints (around 28 and 48 years) where the relationship between age and travel distance shifts direction. This reveals a “gender-masking” effect: aggregate analyses would suggest a negligible gender influence, whereas our approach shows that young women travel farther than young men, but active-caregiving women travel far less than men due to caregiving constraints. Besides, we introduce the notion of a “gender difference amplifier”, quantifying how certain destinations disproportionately extend one gender’s travel over the other. For instance, shopping facilities significantly amplify women’s travel distance during caregiving years, while cultural dining traditions pull post-intensive-caregiving men’s travel farther. These nuanced insights—such as caregiving obligations curbing women’s mobility and social traction drawing post-intensive-caregiving men – were hard to discern with earlier data or methods. By combining big data with interpretable machine learning, this study provides new theoretical and empirical insights into how gendered life-course roles and urban opportunity structures interact to shape everyday non-work mobility.