Corresponding author: Jie He

Abstract

During the transition from experiential to quantitative navigation, the compilation of maritime rutters, exemplified by Shun Feng Xiang Song 順風相送, illustrates a paradigm shift in ancient Chinese traditional navigation technology. Consequently, human practices in maritime spaces were extended, inherited, and replicated across spatiotemporal scales beyond individual cognition. This study addresses the question: how did traditional Chinese navigation technology shape the rules and knowledge structures of human spatial movement across the seas? Innovatively integrating digital humanities with Geographic Information System (GIS) spatial analysis methods, the research quantitatively and structurally extracts descriptions of navigation techniques from historical texts to reconstruct maritime movement strategies/rules under varying environmental conditions. These strategies are then used as foundational logic for GIS spatial analysis to delineate and map the spatial boundaries of ancient maritime navigation.

The methodology encompasses five aspects: digitization, vectorization, structuring, knowledge formalization, and visualization. (1) First, using the Zhenlubu 针路簿 as the source material, the text is scanned and digitized. (2) Second, a locally deployed large language model, incorporating the Qwen visual model, performs high-accuracy OCR on ancient texts, supplemented by manual verification for text proofreading, punctuation correction, and conversion between traditional and simplified Chinese. Systematic prompt engineering used for translation. (3) Third, specialized vocabularies—spatial (including place names, directions, distances), temporal (such as time periods, months, seasons), and environmental (covering water depth, reefs, weather changes, tides, celestial angles)—are established to build a relational database. Using spatial references from the texts, real geographic coordinates are assigned to the spatial vocabulary, achieving georeferencing of historical place names in GIS. (4) Fourth, the relational database organizes and interconnects data via explicit indices, forming the project’s underlying database. Drawing on knowledge graph theory and the CIDOC Conceptual Reference Model (CRM) for cultural heritage, a maritime movement ontology model is constructed to Maritime Movement Strategies Based on Environmental Conditions. (5) Fifth, Large Language Models (LLM) like Gemini assist in developing a web platform for querying navigation routes and interactive map visualization. This methodology constructs a conceptual model that bridges the humanistic ocean and spatial ocean, ultimately enabling the model’s projection into geographic space. Thus, it provides a pathway for translating conceptual understandings of maritime cultural value into tangible geographic entities.

Space serves as a medium for human cognition and practice. Technology not only participates in knowledge production and carries humanistic values but also implies the limits and boundaries of human spatial reach. The mapping of technology onto space holds potential for interpreting the logic of human–ocean interaction and the value of cultural landscapes.