Ontology Integration for Cultural Landscape Management Using ML and Assistive Artificial Intelligence

This paper introduces an automatic ontology model for rare domains, with a focus on cultural landscape management, characterized by internationalization and inter-country relationships. The model fully automates ontology generation by extracting categories and informative terms from the dataset, applying topic modeling, and enriching metadata using generated topics. It strengthens knowledge by incorporating glossary indexes from e-books and leveraging Wikidata. Categorization is achieved through a deep-learning GRU classifier. Semantic relevance is computed using NPMI, PMI, and Petraitis index with differential thresholds, enhancing reasoning infrastructure. The model achieves a high precision of 95.06%, a low False Discovery Rate of 0.05, and an impressive F-measure of 96.06%. This research makes a significant contribution to automatic ontology conception, particularly in rare domains like cultural landscape management.