# NARA: Anchor-Conditioned Representation Learning for Heterogeneous Vector Geoentities

arXiv: https://arxiv.org/abs/2605.12276 · HTML: https://mlsanity.com/paper/2605.12276/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 45 references that are in the map.
- First posted on arXiv 12 May 2026.

## Builds on (most cited first, 21 of 21)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [node2vec: Scalable Feature Learning for Networks](https://mlsanity.com/paper/1607.00653/) (2016)
- [SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery](https://mlsanity.com/paper/2311.17179/) (2023)
- [Tile2Vec: Unsupervised representation learning for spatially distributed data](https://mlsanity.com/paper/1805.02855/) (2018)
- [Multi-Scale Representation Learning for Spatial Feature Distributions using Grid Cells](https://mlsanity.com/paper/2003.00824/) (2020)
- [GeoLM: Empowering Language Models for Geospatially Grounded Language Understanding](https://mlsanity.com/paper/2310.14478/) (2023)
- [SpaBERT: A Pretrained Language Model from Geographic Data for Geo-Entity Representation](https://mlsanity.com/paper/2210.12213/) (2022)
- [Towards General-Purpose Representation Learning of Polygonal Geometries](https://mlsanity.com/paper/2209.15458/) (2022)
- [City Foundation Models for Learning General Purpose Representations from OpenStreetMap](https://mlsanity.com/paper/2310.00583/) (2023)
- [Relational Fusion Networks: Graph Convolutional Networks for Road Networks](https://mlsanity.com/paper/2006.09030/) (2020)
- [Poly2Vec: Polymorphic Fourier-Based Encoding of Geospatial Objects for GeoAI Applications](https://mlsanity.com/paper/2408.14806/) (2024)
- [GeoVectors: A Linked Open Corpus of OpenStreetMap Embeddings on World Scale](https://mlsanity.com/paper/2108.13092/) (2021)
- [Road Network Representation Learning: A Dual Graph based Approach](https://mlsanity.com/paper/2304.07298/) (2023)
- [Road Network Representation Learning with the Third Law of Geography](https://mlsanity.com/paper/2406.04038/) (2024)
- [POI-Enhancer: An LLM-based Semantic Enhancement Framework for POI Representation Learning](https://mlsanity.com/paper/2502.10038/) (2025)
- [GAIR: Location-Aware Self-Supervised Contrastive Pre-Training with Geo-Aligned Implicit Representations](https://mlsanity.com/paper/2503.16683/) (2025)
- [Self-Supervised Representation Learning for Geospatial Objects: A Survey](https://mlsanity.com/paper/2408.12133/) (2024)
- [Geo2Vec: Shape- and Distance-Aware Neural Representation of Geospatial Entities](https://mlsanity.com/paper/2508.19305/) (2025)
- [Building Autocorrelation-Aware Representations for Fine-Scale Spatiotemporal Prediction](https://mlsanity.com/paper/2112.05313/) (2021)
- [MC-GTA: Metric-Constrained Model-Based Clustering using Goodness-of-fit Tests with Autocorrelations](https://mlsanity.com/paper/2405.18395/) (2024)

## Abstract (arXiv)

Vector geospatial data represent the world as discrete geoentities, such as roads, buildings, and points of interest, each with semantic attributes, geometry, and spatial relations to other geoentities, including metric proximity and topology. Existing methods for learning geoentity representations typically support a single geometry type or model only a subset of these relations, limiting their ability to capture spatial context across heterogeneous geoentities and support diverse downstream tasks. We propose NARA (Neural Anchor-conditioned Relation-Aware representation learning), a novel self-supervised representation framework for heterogeneous vector geoentities. NARA contextualizes geoentities through spatial-context-aware attention that models spatial autocorrelation using geometry distance modulated by topological relations across surrounding points, polylines, and polygons. NARA introduces masked geoentity semantic modeling and geometry-aware spatial relation modeling, as well as relation-conditioned regularization that encourages similar representations for geoentities sharing the same spatial relation to a common reference entity, while accounting for spatial autocorrelation. NARA's frozen, task-agnostic encoder outperforms state-of-the-art methods, each with an architecture tailored to its respective task, across traffic-speed prediction for polylines, building-function classification for polygons, and next point-of-interest prediction for points.