Semantic and ontology-based analysis of regulatory documents for construction industry digitalization

In the context of the digitalization of the construction industry, the demand for automation in the analysis of regulatory documents is increasing. The significant volume, structural complexity, and frequent amendments of regulatory acts lead to semantic inconsistencies, duplication of provisions, and contradictions in requirements. The aim of this study is to develop a method for the automated analysis of regulatory documents in the construction sector based on the integration of ontological modeling and natural language processing (NLP) techniques. The relevance of the topic is driven by the need for the digital transformation of construction standardization, which involves eliminating redundant provisions, logical inconsistencies, and outdated references in the regulatory framework. The study proposes a methodology for constructing semantic “profiles” of regulatory statements, which include structured components: subject, predicate, object, modality, and additional conditions. A software prototype has been developed, implementing an algorithm for semantic matching of regulatory requirements using an ontological model that incorporates SKOS-based terminology, deontic logic, and domain-specific concepts of the construction sector. Experiments were conducted on a corpus of 14 regulatory documents of the Republic of Kazakhstan (≈242,000 words), demonstrating high computational efficiency (document analysis time <10 s) and acceptable quality metrics (F1-score up to 0.86). The results confirm the applicability of the proposed method for integration into automated regulatory compliance control systems and Building Information Modeling (BIM).