Scholarlink

<p>Traditional spatial-temporal studies of heritage often rely on dynastic periodization, which may obscure nuanced temporal dynamics. Additionally, there is a lack of interpretability analysis regarding the driving factors behind the patterns. The study focuses on the Sichuan Diocese, subdividing the period from 1640 to 1949 into 26 decadal slices. We introduce a novel “density-weighted historical event segmentation method,” integrating quantitative and historical perspectives to identify optimal chronological divisions, thus providing a robust temporal framework for spatial analysis. To enhance interpretability, a Random Forest model with SHAP analysis is applied to examine the driving factors behind Catholic church distribution. Key findings include: (1) Four distinct stages in church construction: Emergence, Struggling, Rapid, and Golden. (2) Spatial patterns evolved from a Nascent Bipolar to a Networked Pattern. (3) Economic factors had the highest explanatory power, with the interaction between economic and ethnic factors showing the strongest joint influence.</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

Spatiotemporal dynamics of Catholic church heritage in Sichuan using density weighted historical segmentation and SHAP

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月22日

查看全文
  • Huizhen Zhang,
  • Xinkai Li,
  • BeiBei Zhang,
  • Yunzhang Li,
  • Hongyuan Liu

摘要

Traditional spatial-temporal studies of heritage often rely on dynastic periodization, which may obscure nuanced temporal dynamics. Additionally, there is a lack of interpretability analysis regarding the driving factors behind the patterns. The study focuses on the Sichuan Diocese, subdividing the period from 1640 to 1949 into 26 decadal slices. We introduce a novel “density-weighted historical event segmentation method,” integrating quantitative and historical perspectives to identify optimal chronological divisions, thus providing a robust temporal framework for spatial analysis. To enhance interpretability, a Random Forest model with SHAP analysis is applied to examine the driving factors behind Catholic church distribution. Key findings include: (1) Four distinct stages in church construction: Emergence, Struggling, Rapid, and Golden. (2) Spatial patterns evolved from a Nascent Bipolar to a Networked Pattern. (3) Economic factors had the highest explanatory power, with the interaction between economic and ethnic factors showing the strongest joint influence.