Intelligent Processing of Spatiotemporal Data: Developments and Challenges in Historical Geography

Historical geography studies human-environment relations and their spatial evolution in historical periods. Its source materials are characterized by spatiotemporal complexity, semantic instability, and multi-source heterogeneity. With the development of geographic information system (GIS), spatiotemporal databases, and artificial intelligence, historical geography research is gradually shifting from source digitization to data structuring, spatial computation, and intelligent analysis. From the perspective of computer science, this article examines the structural features of historical geographic data, including long-term evolution, place-name disambiguation, multi-source integration, relational modeling, and uncertainty representation. Taking Chinese Historical Geographic Information System (CHGIS) and the China Historical Geographic Information Platform as examples, it further analyzes the implementation paths of spatiotemporal data modeling and spatial analysis in historical geography. On this basis, the article discusses the potential applications of artificial intelligence methods, such as large language models, computer vision, and multimodal learning, in historical document information extraction, old map interpretation, remote sensing image processing, and cross-modal data integration. The study argues that artificial intelligence can enhance the processing and spatial analysis of historical geographic data, but it still has limitations in historical contextual understanding and interpretive reasoning. Future historical geographic computation should continue to advance data standardization, uncertainty modeling, cross-modal alignment, and spatial reasoning, so as to develop a human-machine collaborative research paradigm that combines data-driven methods with knowledge constraints.