Urban Representation Learning for Fine-grained Economic Mapping: A Semi-supervised Graph-based Approach

TL;DR

This work proposes SemiGTX, an explainable semi-supervised graph learning framework for sectoral economic mapping that introduces a semi-information loss function that combines spatial self-supervision with locally masked supervised regression, enabling more informative and effective region representations.

Abstract

Fine-grained economic mapping through urban representation learning has emerged as a crucial tool for evidence-based economic decisions. While existing methods primarily rely on supervised or unsupervised approaches, they often overlook semi-supervised learning in data-scarce scenarios and lack unified multi-task frameworks for comprehensive sectoral economic analysis. To address these gaps, we propose SemiGTX, an explainable semi-supervised graph learning framework for sectoral economic mapping. The framework is designed with dedicated fusion encoding modules for various geospatial data modalities, seamlessly integrating them into a cohesive graph structure. It introduces a semi-information loss function that combines spatial self-supervision with locally masked supervised regression, enabling more informative and effective region representations. Through multi-task learning, SemiGTX concurrently maps GDP across primary, secondary, and tertiary sectors within a unified model. Extensive experiments conducted in the Pearl River Delta region of China demonstrate the model's superior performance compared to existing methods, achieving R2 scores of 0.93, 0.96, and 0.94 for the primary, secondary and tertiary sectors, respectively. Cross-regional experiments in Beijing and Chengdu further illustrate its generality. Systematic analysis reveals how different data modalities influence model predictions, enhancing explainability while providing valuable insights for regional development planning. This representation learning framework advances regional economic monitoring through diverse urban data integration, providing a robust foundation for precise economic forecasting.

Figures (10)

• Figure 1: Overview of the SemiGTX Framework. Multiple data sources are integrated and transformed into a graph structure via specialized pre-encoders. Input data flows through P/S encoder and multiple GraphGPS layers, trained with semi-info loss, to generate multi-task outputs including GDP across primary, secondary, and tertiary economic sectors. SHAP-based interpretation techniques are also applied.
• Figure 1: Administrative Division Schematic of the Pearl River Delta
• Figure 2: Structure of the fine-tuned ViT pre-encoder for SVIs. Each SVI is processed through the ViT encoder to generate its feature vector. During training, the ViT decoder reconstructs the image, and the model is trained by minimizing the discrepancy between the original and reconstructed images.
• Figure 2: Spatial Heatmap of Origin Frequencies for Human Mobility in the Pearl River Delta
• Figure 3: The workflow of the Semi-info loss. Positive sample features are transformed into negative samples via a corruption function. Subgraph-level representations are derived through aggregation and integrated with supervision signals via an MLP predictor to form regression loss. Additionally, positive and negative samples along with subgraph representations are used for self-supervised learning via the graph infomax method. The overall loss function balances these components using a weighting factor.