Tracing Footsteps of Similar Cities: Modeling Urban Economic Vitality with Dynamic Inter-City Graph Embeddings

TL;DR

This work addresses predicting urban economic vitality by learning city embeddings from a dynamic, multi-graph inter-city framework. It introduces ECO-GROW, combining Dynamic Top-K GCN (DTKGCN) and a Graph Scorer to fuse six static networks with temporal industry dynamics, guided by Barabasi Proximity for link prediction. The approach jointly optimizes node growth prediction and inter-city proximity, achieving superior accuracy over baselines and demonstrating robustness across tasks and years. The framework offers practical potential for urban planning and policy, and the authors release open-source code to foster broader applications.

Abstract

Urban economic vitality is a crucial indicator of a city's long-term growth potential, comprising key metrics such as the annual number of new companies and the population employed. However, modeling urban economic vitality remains challenging. This study develops ECO-GROW, a multi-graph framework modeling China's inter-city networks (2005-2021) to generate urban embeddings that model urban economic vitality. Traditional approaches relying on static city-level aggregates fail to capture a fundamental dynamic: the developmental trajectory of one city today may mirror that of its structurally similar counterparts tomorrow. ECO-GROW overcomes this limitation by integrating industrial linkages, POI similarities, migration similarities and temporal network evolution over 15 years. The framework combines a Dynamic Top-K GCN to adaptively select influential inter-city connections and an adaptive Graph Scorer mechanism to dynamically weight cross-regional impacts. Additionally, the model incorporates a link prediction task based on Barabasi Proximity, optimizing the graph representation. Experimental results demonstrate ECO-GROW's superior accuracy in predicting entrepreneurial activities and employment trends compared to conventional models. By open-sourcing our code, we enable government agencies and public sector organizations to leverage big data analytics for evidence-based urban planning, economic policy formulation, and resource allocation decisions that benefit society at large.

Figures (4)

• Figure 1: The maps illustrate industry similarity dynamics and differences in city similarity networks, focusing on a part of China. The star marks Chongqing and the circles indicate the top five similar cities. (a) Shows changes in industry similarity over time. (b) Compares the top five similar cities across attributes.
• Figure 2: (a) The overall architecture of ECO-GROW. (b) DTKGCN Module: For each node in a certain graph $g$, the top-$k_g$ neighbors are selected, retaining the maximum value in each dimension to form the new embedding.
• Figure 3: KL Divergence between years for two downstream tasks. For year $t$, the result represents the distribution shift for each task from the previous year $t-1$ to the current year $t$.
• Figure 4: Sensitivity Analysis of the parameters $\lambda$ and $d$ in predicting the number of new companies respectively.