Intelligent Identification and Correction of Property-related Design Defects in BIM via Large Language Models
Published in Automation in Construction, 2026
This paper proposes an integrated framework to identify and correct multi-type property-related design defects in BIM via Large Language Models, which outperforms rule checking, and the fine-tuned LLM improves identification accuracy by 15%, with a 94% rationality rate for correction suggestions. Proposed hallucination control method increases accuracy from 64% to 85%, eliminating 92.5% of hallucinations in one round
Recommended citation: Lin, J.R., Cai, Y., Ni, X.R., Pan, P.* (2026). Intelligent Identification and Correction of Property-related Design Defects in BIM via Large Language Models. Automation in Construction, 192, 107218. doi: 10.1016/j.autcon.2026.107218 http://doi.org/10.1016/j.autcon.2026.107218