Planning for Subjective Well-Being Using Social Media Data: Nonlinear Effects of Accessibility on Public Sentiment

Read full text Summarise with VIRAT
Authors: Yuanzhao Wang
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

How should planners interpret the relationship between the built environment and subjective well-being (SWB) when effects are nonlinear and context-dependent? Using geotagged Sina Weibo posts from Jiaxing, China (2018), this study derives a proxy for SWB from text-based sentiment analysis and links each post to accessibility measures for eight everyday amenities, using both distance and travel time indicators. Gradient Boosting Decision Tree (GBDT) models, combined with SHAP and partial dependence analyses, are employed to uncover nonlinear association and threshold effects between accessibility and public sentiment. Results show pronounced threshold effects with direct planning relevance: Proximity to green space is associated with higher sentiment at neighborhood-scale distances (within 1km), while sentiment score declines sharply where green space is distant (beyond 5km), underscore the importance of nearby green provision. In contrast, extremely close proximity to bus stops and certain commercial destinations are associated with lower sentiment, suggesting that localized disamenities such as noise, congestion, and crowding can offset accessibility benefits. Distance- and time-based measures are not interchangeable, as they capture different planning dimensions related to walkable local access versus network performance. Limitations include partial representativeness of Weibo users, single year data for single-city design, which constrains generalizability and causal interpretation, and restricted interpretability inherent in GBDT models.

Fields of Study & Subjects

Urban Transport and Accessibility Urban Green Space and Health Psychological Well-being and Life Satisfaction

Publication details

Journal / source
University College London
Publication date
19 March 2026
Language
EN
Document type
Dataset