Leveraging Multi-Source Textural UGC for Neighbourhood Housing Quality Assessment: A GPT-Enhanced Framework

TL;DR

This paper presents a GPT-4o based framework that utilizes multi-source user-generated content to assess neighbourhood housing quality, achieving high accuracy and revealing platform-specific insights for urban planning.

Contribution

It introduces a novel GPT-4o driven method for analyzing multi-source UGC to evaluate housing quality, filling gaps in objective-subjective assessment and platform focus differences.

Findings

GPT-4o outperformed rule-based and BERT models with 92.5% accuracy.

Developed a comprehensive 46-indicator, 11-category assessment system.

Revealed platform-specific differences in UGC focus.

Abstract

This study leverages GPT-4o to assess neighbourhood housing quality using multi-source textural user-generated content (UGC) from Dianping, Weibo, and the Government Message Board. The analysis involves filtering relevant texts, extracting structured evaluation units, and conducting sentiment scoring. A refined housing quality assessment system with 46 indicators across 11 categories was developed, highlighting an objective-subjective method gap and platform-specific differences in focus. GPT-4o outperformed rule-based and BERT models, achieving 92.5% accuracy in fine-tuned settings. The findings underscore the value of integrating UGC and GPT-driven analysis for scalable, resident-centric urban assessments, offering practical insights for policymakers and urban planners.