Mobility-Embedded POIs: Learning What A Place Is and How It Is Used from Human Movement

2026-01-29

AI Summary

This work addresses the limitation of existing geospatial representation methods in jointly modeling both the identity and function of places, particularly the absence of function-aware representations grounded in real human activity. To this end, the authors propose ME-POIs, a novel framework that treats point-of-interest (POI) functionality as a core signal. ME-POIs aligns large-scale human mobility trajectories with POI textual embeddings via contrastive learning and introduces a multi-scale spatiotemporal visitation pattern propagation mechanism over neighboring POIs to effectively mitigate long-tail sparsity. The resulting representations are POI-centric, context-independent, and reflective of actual usage patterns. Evaluated on five newly introduced map-enhancement tasks, ME-POIs significantly outperforms baselines relying solely on text or mobility data; notably, models trained exclusively on mobility data surpass text-based models on several tasks, underscoring the critical value of explicit POI functional modeling.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Recent progress in geospatial foundation models highlights the importance of learning general-purpose representations for real-world locations, particularly points-of-interest (POIs) where human activity concentrates. Existing approaches, however, focus primarily on place identity derived from static textual metadata, or learn representations tied to trajectory context, which capture movement regularities rather than how places are actually used (i.e., POI's function). We argue that POI function is a missing but essential signal for general POI representations. We introduce Mobility-Embedded POIs (ME-POIs), a framework that augments POI embeddings derived, from language models with large-scale human mobility data to learn POI-centric, context-independent representations grounded in real-world usage. ME-POIs encodes individual visits as temporally contextualized embeddings and aligns them with learnable POI representations via contrastive learning to capture usage patterns across users and time. To address long-tail sparsity, we propose a novel mechanism that propagates temporal visit patterns from nearby, frequently visited POIs across multiple spatial scales. We evaluate ME-POIs on five newly proposed map enrichment tasks, testing its ability to capture both the identity and function of POIs. Across all tasks, augmenting text-based embeddings with ME-POIs consistently outperforms both text-only and mobility-only baselines. Notably, ME-POIs trained on mobility data alone can surpass text-only models on certain tasks, highlighting that POI function is a critical component of accurate and generalizable POI representations.

Problem

• POI representation
• human mobility
• place function
• geospatial foundation models
• map enrichment

Innovation

• Mobility-Embedded POIs
• POI function
• contrastive learning
• temporal visit patterns
• spatial propagation