跳至正文
Home » Projects » Public sentiment & Urban Complex System

Public sentiment & Urban Complex System

Urban Complex System Impacts on Public Sentiment under Public Health Emergencies : A Causal Inference Study based on ChatGPT & Social Media Data

Motivation

Our research aims to identify whether neighbourhood liveability influences mental health, especially under circumstances of health crsis.

Introduction

The public health crisis not only affects the physical health of the public, but also the control measures such as closure, traffic restriction and production suspension, which will have a negative impact on the mental health of the public, thus triggering the turmoil of the social and economic system. Therefore, it is necessary to study the temporal and spatial characteristics and influencing mechanism of public emotion under public health crisis. This project takes Shanghai as the research area, integrates multiple big data such as microblog, remote sensing image and point of interest, and applies chatGPT, mood map, deep learning and other means to conduct quantitative research on the temporal and spatial evolution and influence mechanism of public emotions under the full cycle of public health crisis.


Firstly, we collected and cleaned over 900,000 Weibo check-in data in Shanghai since 2021, introduced chatGPT to realize text semantic analysis, extracted social media discussion topics, classified emotions and calculated intensity values, and realized dynamic monitoring of public opinion. Secondly, the original Geographically weighted artificial neural network (GWANN) model was adopted to solve the geographically weighted artificial neural network (GwanN) problem. The influence mechanism of built environment, blue and green space, social and economic factors on public sentiment is analyzed, and relevant planning and policy suggestions are proposed based on the research conclusions.


In terms of space planning practice, this study can guide the layout of 15-minute living circle facilities, the improvement of blue and green space environment and other planning and construction measures. At the level of city management, it can guide real-time public opinion response and formulation of health crisis control policies. This study innovatively applied chatGPT to perform emotion analysis and original GWANN model, which can be extended to public opinion emergency response and planning management of various disasters in the city, helping to improve the resilience level of urban health.

Figure. Public Sentiment (Blue line for 2021 and orange line for 2022)

Coming soon

发表回复

您的电子邮箱地址不会被公开。 必填项已用 * 标注