Mingshu’s Group Presents Latest Research at the First International GeoAI Conference

Dr Mingshu Wang and doctoral students Xinyi Yuan and Rui Deng recently attended the first International GeoAI Conference in Ghent, Belgium (2–6 June 2026). Both PhD students delivered 15-minute oral presentations on their recently published research.

Xinyi Yuan presented her work, entitled ‘Where England’s cities are growing: Evidence from big building footprint data and explainable AI’ (DOIs: https://doi.org/10.1177/23998083251317573 and https://doi.org/10.1016/j.habitatint.2025.103457). Her research utilises massive building footprint datasets and Explainable AI (XGBoost and SHAP) to model urban structural changes across England from 2017 to 2023. The study highlights the impacts of population density, the Index of Multiple Deprivation (IMD), and regional demographics on urban development, synthesising findings from her recent paper in Habitat International and Environment and Planning B.

Rui Deng introduced ‘GeoAggregator: An Efficient Transformer Model for Geospatial Tabular Data’ (DOI: https://doi.org/10.1609/aaai.v39i11.33259). He detailed the GeoAggregator framework, designed to effectively address spatial dependence and heterogeneity in geospatial tabular datasets. A key highlight was the release of the open-source GA-Sklearn Python package, which provides researchers with an accessible and highly efficient tool to process and model their own geospatial data

This milestone event was a fantastic opportunity to connect with international peers in the rapidly advancing GeoAI field. We look forward to more exciting collaborations and sharing further updates with the community in the future!