GeoAI in social science

This chapter introduces GeoAI, an emerging field that integrates artificial intelligence, geospatial big data, and high-performance computing for geospatial problem solving. It starts with presenting the unique opportunity GeoAI offers for deepening our understanding of the social systems by serving as an advanced spatial-social analytical technique of big data. Next, the chapter introduces two main threads of GeoAI research methods: the bottom-up data-driven approach, represented by deep learning; and the top-down ontological approach, exemplified by knowledge graph. Two social science use cases are then introduced to demonstrate the applicability and potential of GeoAI approaches to be seamlessly integrated into the social science research framework. These are (1) the application of deep learning for estimating social demographic information at the neighborhood scale; and (2) a disease knowledge graph for spatial and temporal question answering about COVID-19 outbreak. The chapter concludes with a discussion of the remaining challenges and future research directions of GeoAI in social science.