Cyber-GIS Center for Advanced Digital and Spatial Studies

🌟 I-GUIDE Platform Knowledge Element of the Week!

For the week of August 3, "Do Location Encoders Capture Spatial Effects? A GeoShapley Benchmark Across Scales" by Daniel Kiv and Shaowen Wang is the knowledge element of the week.

Benchmark framework for evaluating how well spatial location encoders recover spatially-varying coefficients (SVCs) from synthetic data, using GeoShapley for post-hoc spatial effect extraction.

✨ Access the element here: https://platform.i-guide.io/code/2d2fb19e-b082-4e1e-99e9-a83d9148e3c4

The 2026 I-GUIDE Forum and Harnessing the Revolution (HDR) Ecosystem Conference begins today at the University of Illinois Chicago! 🏙️

🌍 Running through today until Friday, August 7, this year’s conference explores the theme & : From Geospatial to Convergence through workshops, tutorials, keynote and plenary presentations, panels, research talks, posters, and collaborative discussions.

🎤 Attendees will hear from keynote speakers Luc Anselin, Chaitan Baru, and Shaowen Wang, as well as leaders from four HDR Institutes. Sessions will explore timely topics including Spatial AI, and applications, responsible and trustworthy AI, AI literacy and workforce development, convergence science, and disaster resilience.

Join us for an exciting week of learning, conversation, and collaboration!

🔗 Register today: https://i-guide.io/forum/forum-2026/forum-2026-and-hdr-ecosystem-conference-registration/

🏙️ Spatial AI for Micro-scale Urban Accessibility Modeling

Join Rafael Albuquerque, Jessica Miranda, and Vinicius Andrade Brei of the Federal University of Rio Grande do Sul for an upcoming VCO featuring the 3rd Place project from the I-GUIDE Spatial AI Challenge 2025–26.

Learn how the GeoSocial Downscaling Model (GSDM) uses Spatial AI and physics-guided deep learning to reconstruct neighborhood-scale accessibility from coarse human mobility data. The session will explore how high-resolution accessibility mapping can support resilience planning, environmental justice, and equitable access to services through reproducible geospatial AI workflows.

📅 August 19, 2026 | 🕚 11:00 am (Central Time)
🔗 https://i-guide.io/i-guide-vco/spatial-ai-for-micro-scale-urban-accessibility-modeling/

🌟 New publication from the NSF I-GUIDE Team!

This publication is from Environmental Research Letters, "Regional drought impacts drive telecoupled land change through international agricultural trade" by Nicholas Manning, Iman Haqiqi, Thomas Hertel and Jianguo Liu.

📃 Read the full paper here 👉 https://iopscience.iop.org/article/10.1088/1748-9326/ae81c9

📢 Registration is now open for the I-GUIDE Forum 2026 & HDR Ecosystem Conference!

Join researchers, educators, industry professionals, and policymakers in Chicago from August 3–7, 2026, to explore the intersection of AI and science—from geospatial discovery and Spatial AI to convergence research, CyberGIS, and workforce development.

This year’s joint event brings together the I-GUIDE Forum and the HDR Ecosystem Conference, creating a unique opportunity to connect across disciplines, engage with cutting-edge research, participate in hands-on workshops and tutorials, and help shape the future of data-intensive science.

📍 University of Illinois Chicago
📅 August 3–7, 2026
🎓 Student registration available
🔗 Register and learn more: https://i-guide.io/forum/forum-2026/

🌟 I-GUIDE Platform Knowledge Element of the Week!

For the week of July 27, "RegionLM - Notebook Tutorial" by Yao-Yi Chiang, Yijun Lin, and Zekun Li is the knowledge element of the week.

RegionLM is a geospatial representation learning pipeline built around SpaBERT-style contextual embeddings for OpenStreetMap (OSM) features. It extracts features inside target regions, converts nearby spatial context into pseudo-sentences, trains or applies a spatial BERT model, aggregates feature embeddings into region embeddings, and clusters the resulting regions. The repository is organized as a script-driven research workflow. This notebook walks through the intended end-to-end sequence, while the numbered Python scripts provide CLI entrypoints for each stage. Check out the Open Educational Resource that comes with this Notebook too!

✨ Access the element here: https://platform.i-guide.io/notebooks/8e5b10d1-e2a1-4386-a7e9-e4b31810fcd0

📘Open Educational Resource: https://platform.i-guide.io/oers/74e05c45-dbb0-4b05-a91d-07c26dc70a76

Last week, participants from across the country came together at the I-GUIDE Summer School 2026 to explore how Spatial AI, data science, and convergence can address real-world challenges.

Throughout the week, interdisciplinary teams collaborated on innovative projects spanning wildfire mitigation, urban heat, remote sensing, environmental monitoring, , and more—gaining hands-on experience while building solutions with real-world impact.

Explore this year's Summer School projects: https://i-guide.io/summer-school/summer-school-2026/summer-school-2026-projects/
I-GUIDE platform: https://platform.i-guide.io/

🌟 I-GUIDE Platform Knowledge Element of the Week!

For the week of July 20, "GeoLocate: Spatial Modeling of Market Entry Viability" by Jaiany Rocha Trindade, Devika Jain, and Vinicius Andrade Brei is the knowledge element of the week.

Market entry is highly uncertain, with non-spatial methods often ignoring spatial dependence and providing limited support for geographic decision-making. This study develops a Bayesian Spatial AI model using a Besag-York-Mollié 2 (BYM2) framework and Bayesian Lasso regularization to predict business survival probabilities. The model integrates data from OpenStreetMap, Sentinel-2 imagery, IBGE census data, and the Brazilian Federal Revenue and is validated across the Retail and Food & Beverage sectors in São Paulo and Rio Grande do Sul. The results show that spatial dependence accounts for 85% of the residual variation. Bayesian Lasso identified income, accessibility, and urbanization as key predictors. Compared to non-spatial models, this approach significantly improves predictive performance (elpd_diff = 28.48), eliminates residual autocorrelation, and provides geographically explicit uncertainty quantification, enabling risk-aware spatial decision-making for entrepreneurs and policymakers.

✨ Access the element here: https://platform.i-guide.io/notebooks/cd862a29-0707-4bd0-8e7c-a07a62710c1b

🌐 Missed our recent I-GUIDE VCO? The recording is now available!

Thank you to Yao-Yi Chiang (University of Minnesota), Yijun Lin (Southern Illinois University, Carbondale), and Zekun Li (University of Minnesota) for an insightful session on RegionLM: Integrating Point, Line, and Polygon Context for Urban Representation Learning.

The presentation introduced RegionLM, a Spatial AI framework that learns contextual representations of urban regions from OpenStreetMap data. The session explored how semantic, spatial, and topological information can be combined to support applications such as mobility modeling, neighborhood analysis, and trajectory anomaly detection.

🎥 Watch the recording here: https://www.youtube.com/watch?v=OGaGPCcyOlE

RegionLM: Integrating Point, Line, and Polygon Context for Urban Representation Learning RegionLM is a geospatial representation learning pipeline for gener...