Location encoders learn boundaries and place names from reverse geocoding
Learning Administrative Boundary-Aware Location Encoders via Reverse Geocoding
Machine LearningArtificial Intelligence
Summary
Representing locations as numbers is common in mapping and prediction tasks, but existing methods often miss important administrative boundaries like city or state borders. The authors developed a way to teach location models using the text output of reverse geocoding (which translates coordinates into place names) as a training signal. This method lets models better understand and respect these boundaries, improving their accuracy on tasks involving location data. They also created a specialized model that performs even better and produces location codes people can understand.
What this means in practice
- •For geospatial software developers: Improve geographic prediction models by integrating administrative boundary-aware location representations derived from reverse-geocoded text.
- •For urban planners: Use enhanced location encodings to better model and analyze urban phenomena influenced by administrative boundaries and place names.
- •For marketing analysts: Create more precise market segmentation models that recognize administrative boundaries using interpretable location representations combining coordinates with place names.$Commercial implications: Enables development of location-aware marketing tools that target customers more accurately by administrative regions using novel location encoding methods.
Authors
Lakshay Sharma, Alex Marin
Abstract
Location encoders learn compact representations of latitude--longitude coordinates for downstream geospatial prediction tasks. Existing approaches are pretrained using continuous or naturally occurring supervision, such as imagery, climate variables, or species distributions, and therefore do not explicitly capture administrative boundaries or place names, despite their strong influence on many real-world geospatial phenomena. We introduce, to our knowledge, the first method for learning administrative boundary-aware location encoders, using reverse-geocoded text as a self-supervised training signal. Our approach can be applied to existing location encoder architectures and consistently improves their downstream performance. We further introduce RevGeoRQ, a custom architecture tailored to reverse-geocoding supervision, which yields additional gains and exceeds contemporary location encoders on several downstream tasks. Finally, the use of human-interpretable reverse-geocoded strings and residual-quantized semantic IDs provides a more interpretable representation of geographic location than existing approaches. Code and model weights to be released soon.