Knowledge‐Based <scp>GIS</scp> : Bridging Domain Expertise and Geographic Information Science

ABSTRACT Over the past two decades, technological advances in areas such as big data, artificial intelligence (AI), and digital twin have dramatically reshaped geographic information systems (GIS) and their underlying science. Amidst these shifts, the integration of domain knowledge with GIS remains imperative, and such integration is now more critical than ever for addressing the complexity of spatial analysis and decision‐making. This special issue examines knowledge‐based GIS (K‐GIS) in the AI era and compiles research on knowledge representation, extraction, and reasoning, together with applications across domains. These studies infuse domain knowledge into geospatial tasks through both symbolic constructs, such as ontologies and knowledge graphs, and sub‐symbolic learning, such as spatial representation learning. We frame K‐GIS not as a new term but as the contemporary, expanded form of knowledge‐based GIS, whose original rule‐based and symbolic connotation has broadened to encompass hybrid knowledge representation and semantic interoperability that links data, knowledge, models, and actions. By weaving together domain expertise, semantic technologies, machine learning, and spatial analysis, K‐GIS reshapes how geographic information is represented, analyzed, and operationalized, while opening opportunities for interdisciplinary collaboration and more intelligent spatial services. We close by discussing the theoretical foundations that underpin K‐GIS and the open challenges that determine how far these capabilities can be realized.