Urban region embeddings improve when factoring out hidden shared causes
When Correlations Mislead: Confounder-Aware Multi-View Urban Region Representation Learning
Machine LearningArtificial Intelligence
Summary
Understanding different parts of a city using many data types like travel patterns and land use can help with planning and safety. The authors noticed that combining these data views often mixes up hidden factors that affect all views, causing less reliable results. They created a new method called CURE that first identifies and removes these hidden shared influences before combining the data. This approach helps produce better predictions and stays reliable even when some data is missing or noisy.
What this means in practice
- •For urban planners: Generate more accurate and robust city region models for predicting mobility and service needs by separating hidden shared influences across data sources.
- •For public safety teams: Improve forecasting of urban safety risks by using confounder-aware multi-source data integration that reduces misleading correlations.
Authors
Sean Bin Yang, Ying Sun, Zongyi Xu, Tung Kieu, Jilin Hu, Bin Yang, Kristian Torp, Hua Lu, Torben Bach Pedersen
Abstract
Urban region representation learning commonly combines heterogeneous data sources, such as mobility flows, points of interest, and land-use information, to support tasks including mobility analysis, public safety forecasting, and service demand estimation. Existing multi-view methods typically improve region embeddings by strengthening interactions across views. However, such methods often overlook view-specific regional structures and may propagate correlations induced by shared latent factors, which can reduce the stability of downstream predictions. To overcome this major limitation, we propose CURE, a confounder-aware framework for multi-view urban region representation learning. CURE first encodes each view with its regional graph structure, estimates a shared latent component, and then reduces its projected influence before cross-view interaction. A hierarchical graph-aware fusion module subsequently aggregates the residual view representations using local and global regional contexts Experiments on three real-world cities show that CURE improves predictive performance, remains robust under missing and noisy input views, and provides reliable cross-view integration through shared component separation and context-dependent view weighting.