2025
Chapter
The paper addresses the problem of lengthy and error-prone preparation of design documentation in the construction industry, caused by the increasing volume and fragmentation of regulatory texts. To overcome this challenge, a methodological approach is proposed that combines a systematic analysis of the Code of Practice (CoP) 20.13330.2016 «Loads and Actions» with ontological modeling techniques and the application of large language models (LLMs) to extract key entities, attributes, and logical interrelations. A two-level ontology was developed, encompassing both universal entities (such as load types, material characteristics, and climatic factors) and detailed requirements of the Code of Practice, represented through formulas, constraints, and conditions of applicability. Based on the extracted structures, a knowledge base was constructed and tested using a case study involving the load calculation for a dome-shaped timber building located in Climate Zone VII. Expert validation confirmed the accuracy and completeness of the generated data. The proposed approach demonstrates the potential to transform fragmented regulatory provisions into a unified machine-readable model that sup-ports intelligent search, automated compliance verification, and digital design processes within the framework of smart regulatory documentation.