A systematic review of methods for interpreting building code regulations in automated compliance systems

Automating the interpretation of building code regulations is critical for enhancing accuracy, efficiency and transparency in construction compliance. However, regulatory complexity, dynamic updates and jurisdictional diversity remain significant barriers. This systematic review critically evaluates state-of-the-art computational approaches, including rule-based reasoning, ontology-driven frameworks, Natural Language Processing (NLP), Machine Learning (ML) and Large Language Models (LLMs). Through a comparative analysis across accuracy, efficiency, reliability and explainability, this study identifies distinct trade-offs: rule-based and ontology-driven systems ensure deterministic logic and legal traceability but lack scalability. Conversely, ML and transformer-based models (e.g. BERT, GPT-4) offer superior adaptability but face challenges in interpretability and cross-jurisdictional generalization. The review suggests that hybrid AI architectures, integrating symbolic transparency with neural adaptability, represent a promising pathway for future systems. The study concludes by outlining a strategic research agenda to bridge current methodological gaps. It specifically emphasizes the necessity for standardized benchmarking datasets, cross-jurisdictional ontology alignment and explainable frameworks to achieve scalable and legally robust automated compliance checking.