Artificial intelligence and machine learning applications (AI/ML) provide unique opportunities for documentation, assessment and protection of world cultural heritage, at risk from natural disasters such as earthquakes, climate change, uncontrolled urbanization, etc. Some examples include the deep knowledge bases of churches/mosques damaged in the earthquakes in Italy in 2016, Mexico 2017, Morocco 2023 and Tukey/Syria 2023. Ontology development provides a unifying view of this data and is crucial for effectively applying deep learning techniques to document and manage heritage structures information. A well-designed ontology serves as a structured knowledge graph that facilitates data integration, search, and analysis. Key aspects of ontology development for heritage structure documentation include Ontology Design to ensure the ontology is generic enough to be applicable across diverse heritage assets, Integration of fragmented data from various sources enabling efficient data acquisition and storage, Deep Learning Applications employing the ontology to support deep learning algorithms for semantic segmentation of point cloud data, automated generation of heritage building information models (HBIM) among others, and Knowledge Completion, through models to predict missing information and enhance the quality of the knowledge graphs. Combining ontology development with deep learning techniques, improves heritage structure documentation significantly in terms of accessibility, interpretation, collaboration, and preservation.