Integrating Natural Language Processing and Building Information Modelling for Automated Information Extraction, Regulatory Compliance, and Quantity Take-Off in Infrastructure and Construction Projects

Abstract

The construction and infrastructure sector is undergoing a profound digital transformation driven by the increasing adoption of Building Information Modelling (BIM) and advances in Natural Language Processing (NLP). Despite these developments, a significant proportion of critical project information remains embedded in unstructured or semi-structured textual documents, such as regulations, contracts, inspection reports, bills of quantities, and work descriptions. This fragmentation limits automation, introduces interpretive ambiguity, and increases the likelihood of errors in regulatory compliance, quantity take-off, cost estimation, and project control. Drawing strictly upon established scholarly works in construction informatics, BIM-enabled quantity take-off, ontology-driven modelling, and NLP-based information extraction, this research develops an integrated conceptual and methodological framework that unifies NLP techniques with BIM-based workflows. The article provides an extensive theoretical elaboration on how rule-based, ontology-based, machine learning, and hybrid deep learning NLP approaches can be systematically aligned with BIM data structures to automate utility permitting, compliance checking, accident cause classification, bridge inspection interpretation, and high-accuracy quantity take-off. Through descriptive methodological synthesis and interpretive analysis of prior validated research, the study demonstrates that the convergence of NLP and BIM not only reduces human-dependent interpretation but also enhances transparency, consistency, and scalability across the project lifecycle. The findings suggest that such integration represents a foundational shift from document-centric to knowledge-centric construction management. Limitations related to data heterogeneity, domain vocabulary evolution, and model transferability are critically discussed, and future research directions are proposed to support the maturation of intelligent, regulation-aware, and cost-reliable construction systems.