Social Media Analysis Research 1

Assessing and interpreting perceived park accessibility, usability and attractiveness through texts and images from social media

Xukai Zhao, Yuxing Lu, Wenwen Huang, and Guangsi Lin*
赵旭凯 陆宇星 黄雯雯 林广思*
Sustainable Cities and Society (2024) (JCR Q1, IF=12.0)
https://doi.org/10.1016/j.scs.2024.105619

This study introduces the Park Dual-modal Perception (PDP) model, which uniquely combines text and image analysis from social media data to evaluate public perceptions of urban parks. Applied to 130 parks in Guangzhou, the model achieved 86.81% accuracy in assessing accessibility, usability, and attractiveness, outperforming standard models.

Urban park Public perception Natural language processing Computer vision Social media data
Social Media Analysis Research 2

Exploring temporal and spatial patterns and nonlinear driving mechanism of park perceptions: A multi-source big data study

Xukai Zhao, He Huang, Guangsi Lin*, and Yuxing Lu
赵旭凯 黄鹤 林广思* 陆宇星
Sustainable Cities and Society (2025) (JCR Q1, IF=12.0)
https://doi.org/10.1016/j.scs.2024.106083

This research uses a fine-tuned Chinese-RoBERTa-wwm-ext model to analyze over 150,000 online comments on 270 parks in Beijing and Guangzhou, assessing perceived accessibility, usability, and attractiveness. The study identifies significant seasonal and spatial patterns in park use and employs machine learning to determine the nonlinear drivers behind public perceptions.

Urban park Perception Social media data Natural language processing Two-step floating catchment area method Explainable machine learning
Social Media Analysis Research 3

How to quantify multidimensional perception of urban parks? Integrating deep learning-based social media data analysis with questionnaire survey methods

Wenwen Huang, Xukai Zhao, Guangsi Lin*, Zhifang Wang, and Mengyun Chen
黄雯雯 赵旭凯 林广思* 王志芳 陈梦芸
Urban Forestry & Urban Greening (2025) (JCR Q1, IF=6.7)
https://doi.org/10.1016/j.ufug.2025.128754

This paper develops a deep learning model based on ERNIE to assess six dimensions of urban park perception—accessibility, safety, aesthetics, attractiveness, maintenance, and usability—using social media data. The model's effectiveness, achieving over 80% accuracy, was validated by comparing its results with a traditional questionnaire survey.

Multidimensional Perception Urban Parks Deep Learning Social Media Analysis Survey Methods

开始试用 Start Trial

我们提供两种预训练模型供您体验。请选择一个模型,然后从随机生成的评论中选择一条进行分析。 We provide two pre-trained models for you to try. Please select a model, then choose one of the randomly selected comments to analyze.

高精度模型High-Accuracy Model
Chinese-RoBERTa-wwm-extDimensions:
可达性Accessibility 可用性Usability 吸引力Attractiveness
多维度模型Multi-Dimension Model
ERNIEDimensions:
美感Aesthetics 安全感Safety 维护管理Maintenance 可达性Accessibility 可用性Usability 吸引力Attractiveness

第二步:选择一条评论开始分析 Step 2: Select a Comment to Analyze

🔄

正在分析中... Analyzing...

各维度情感概率分析Sentiment Probability Analysis by Dimension