Supplementary Materials (S1–S8) for "Automated BIM-Based Regulatory Compliance Checking for Bridge Infrastructure Using Ontology-Driven Semantic Rule Interpretation: A Multi-Model Validation Study (AASHTO LRFD Case Study)"

This dataset contains eight supplementary files (S1–S8) for the manuscript submitted to Automation in Construction (manuscript ID: AUTCON-D-26-02904). S1: SPARQL queries for rules R1–R4 (Apache Jena 4.9). S2: SHACL NodeShape for rule R5 (TopBraid SHACL API v1.4.3, CWA). S3: Bridge Compliance Ontology (BCO) full OWL 2 DL Turtle serialisation (IRI: https://w3id.org/bco/bridge/, CC BY 4.0). S4: Anonymised RDF triple dumps for bridge models M1–M5; all project-identifying information removed; rule-relevant property attributes sufficient to reproduce all F1 and timing results reported in the manuscript. S5: NLP annotation codebook (five-class regulatory sentence taxonomy) and inter-annotator agreement protocol (150-sentence IAA subset, κ=0.81). S6: BERT-base-uncased fine-tuning script with three random seeds (42/123/456); test accuracy 92.9±0.6%; AdamW optimiser; Python 3.11 / transformers 4.35.0. S7: Manual compliance review session instruction sheet and facilitator timing form (25 sessions: 5 reviewers × 5 models). S8: Statistical analysis script reproducing all Wilcoxon signed-rank, BCa bootstrap (10,000 resamples), and Holm–Bonferroni results from Tables 9 and 10; Python 3.11 / SciPy 1.11.