Abstract

The promotion of residents’ positive sentiments is essential for achieving the Sustainable Development Goals. However, there is limited evidence on the effect of the settlement environment (SE) on sentiment, especially in urban villages (UVs). By combining affective geography and social media data, this study aims to analyze the residents’ sentiments in UVs and formal settlements (FSs) in Shenzhen, while exploring the underlying mechanisms of SE variables that influence the positive sentiment index (PSI). The Weibo text data was analyzed using Natural Language Processing to obtain the PSI. Furthermore, we employed an XGBoost model, Shapley Additive Explanations and Partial Dependence Plots to explore relationships between SE variables and the PSI. We utilized the Interpretative Structural Modeling and Bayesian Network to analyze and verify the interdependencies and probabilistic results. The results revealed that the PSI exhibited spatial heterogeneity, with a trend of medium-high-low from central to suburban areas, and a clustering effect of high and low values. For FSs, we recommend enhancing health and well-being by increasing metro facilities, commercial density and fostering walkable neighborhoods. For UVs, prioritizing micro walk accessibility can improve settlement circulation. Additionally, we identified the potential of marginalized UVs to integrate with e-trade and transform into “special economic zones.”

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call