Abstract: With the expansion of indoor space scale and the complexity of its structure, people's demand for navigation in large and complex indoor spaces has become increasingly strong. IndoorGML is a new standard for representing the topological structure of indoor spaces, primarily used for expressing and exchanging indoor space navigation models, providing a framework and rich semantic information for indoor space information exchange. Most research based on IndoorGML revolves around the expression of indoor space models, while research on navigation applications based on IndoorGML is relatively scarce. In order to further expand the research in this area, this paper proposes an indoor semantic navigation method based on the IndoorGML model, achieving topological expression and semantic navigation of indoor spaces, and providing a new approach for semantic navigation in large and complex spaces.