Disaster Response and Community Resilience: Our research in this area uses GeoAI methods and geospatial data to discover knowledge that can help communities become more resilient in the face of natural disasters. Some of our studies include understanding and extracting location descriptions used by people during natural disasters, examining the impacts of disasters on vulnerable communities, and understanding the perceptions of people toward their living environments. Part of our research also seeks to understand how AI models could be biased due to the lack of training data from minority population groups, a topic that has become increasingly important given the wide use of AI.

Public Health and Environment: The places where we live, work, and spend time can directly affect our health outcomes. Our research in this area aims to assess and understand environmental impacts on public health. AI methods integrated with various types of big geo data (e.g., point-of-interest data, GPS locations, environmental sensors, and remote sensing images) provide novel approaches for improving our understanding of the places that have positive impacts on health (e.g., parks) and those that may have negative impacts (e.g., liquor stores). Such an improved understanding will enable more comprehensive answers to health-related questions from a spatial perspective.

Ecosystems and Biodiversity: Ecosystems and biodiversity provide critical resources for our humans to survive and prosper, which include oxygen, food, clean air and water, pest control, and many others. Our research in this area integrates AI methods, big geo data (e.g., remote sensing images and field observations), and spatial models to discover knowledge that advances our understanding of spatial ecological processes (e.g., vegetation dynamics) and informs decisions related to environmental protection and resource management. Our work currently focuses on the Cape Floristic Region of South Africa, which is a global biodiversity hotspot and a UNESCO World Heritage site.

GIS and GeoAI Methods: Basic research in the above three areas inspires us to develop and improve GIS and GeoAI methods and to enhance spatial data infrastructures (SDIs). Improved GIS and GeoAI methods can help us better analyze geospatial data, and enhanced SDIs can facilitate the reuse of existing datasets, tools, and services. Our research in this area has developed methods for prioritizing disaster mapping tasks, enriching the metadata of geospatial resources, and building knowledge graphs from historical maps. We are also one of the participants that developed the five-star guide for achieving replicability and reproducibility in GIScience research.