All Issue

2026 Vol.39, Issue 1 Preview Page

Research Paper

Development of a GraphRAG-Based Automated Knowledge Graph Query Generation Model for Inference Search in BIM Data: A Case Study of a Rigid-Frame Bridge

BIM 데이터의 추론 검색을 위한 GraphRAG기반 지식그래프 질의 자동 생성 모델 개발: 라멘교를 중심으로

Hyunwoo Cha1
,
Inwoo Jung2
,
Hyounseok Moon3*
차 현우1
,
정 인우2
,
문 현석3*
1Student Researcher, Future Smart Construction Research Division, Korea Institute of Civil Engineering and Building Technology (KICT) & University of Science and Technology (UST), Ilsan, 10223, Korea
2Postdoctoral Researcher, Future Smart Construction Research Division, Korea Institute of Civil Engineering and Building Technology (KICT), Ilsan, 10223, Korea
3Research Fellow (Professor), Future Smart Construction Research Division, Korea Institute of Civil Engineering and Building Technology (KICT) & University of Science and Technology (UST), Ilsan, 10223, Korea
1한국건설기술연구원 미래스마트건설연구본부, 과학기술연합대학원대학교(UST) 학생연구원
2한국건설기술연구원 미래스마트건설연구본부 박사후연구원
3한국건설기술연구원 미래스마트건설연구본부, 과학기술연합대학원대학교(UST) 연구위원, 교수
*Corresponding Author
28 February 2026. pp. 59-70
PDF XML Citation
Abstract

Recently, Building Information Modeling (BIM) has evolved beyond three-dimensional modeling, increasingly focusing on the quality of standardized attribute data and systematic data management. Accordingly, ontology- and knowledge-graph–based data representation and reasoning approaches have increasingly gained attention. Knowledge-graph–based BIM data can support practical tasks such as structural relationship analysis, design review, and quantity and attribute information retrieval. However, practical adoption is limited by the need to manually write graph query languages, such as SPARQL and Cypher, which requires detailed knowledge of graph schemas and query syntax. To address this limitation, a GraphRAG-based automated query generation framework that converts natural-language questions into graph queries is proposed. A CSV-to-Labeled Property Graph (LPG) pipeline is developed to automate graph construction by generating nodes and relationships from rule-based attribute and relationship tables derived from bridge BIM data. In addition, a few-shot learning–based prompt design is utilized to generate Cypher queries from natural-language inputs automatically. The proposed framework is implemented as the Graph-ACQ system and validated using rigid-frame bridge BIM data. Experimental results demonstrate 100% accuracy in schema validity and query generation, with an average processing time of 7.1s. The proposed approach enables intuitive, relationship-oriented BIM data querying and supports scalable application across diverse infrastructure domains.

Keywords
BIM
Knowledge Graph
GraphRAG
Labeled Property Graph
Large Language Model
Few-shot Learing

최근 BIM은 단순 3차원 모델링을 넘어 표준화된 속성 데이터의 품질 확보와 체계적 관리가 핵심 요구로 부각되고 있으며, 온톨로지 및 지식그래프 기반의 데이터 관리･추론 방식이 주목받고 있다. 그러나 지식그래프 기반 BIM 데이터는 부재 간 구조 관계 분석, 설계 검토, 물량･속성 정보 조회 등 실무 의사결정을 지원할 수 있으나, SPARQL･Cypher와 같은 그래프 질의어를 직접 작성해야 한다는 점에서 실무 적용에 제약으로 작용한다. 이를 위해 본 연구에서는 사용자가 자연어 질문을 기반으로 그래프 질의를 자동 생성할 수 있는 GraphRAG 기반 질의 자동 생성 프레임워크를 제안하였다. 먼저 CSV 기반 속성/관계 테이블에 규칙을 적용해 노드･관계를 생성하고 그래프 데이터베이스에 적재하는 CSV-to-LPG 파이프라인을 구현하여, LPG 지식그래프 구축 절차를 자동화하였다. 이후 Few-shot Learning 기반 프롬프트 설계를 통해 사용자의 자연어 질문을 Cypher 쿼리로 자동 변환하는 자동 질의 생성 모듈을 구현하였다. 전체 프레임워크는 Graph-ACQ 시스템으로 개발하여 라멘교 BIM 데이터를 기반으로 적용하였다. 검증 결과 LPG 스키마 유효성과 Cypher 자동 생성, Cypher 질의 수작업 과정에서 정확도 모두 100%를 달성하였고, 질의 생성 시간은 평균 7.1초에 처리되었다. GraphRAG 기반 질의 생성 방식은 부재 간 공간･구조 관계를 명시적으로 활용하므로, 설계･검토 과정에서 요구되는 연결 관계 분석, 구조 구성 파악, 물량･속성 정보 조회 등 관계 기반 질의를 자연어로 수행할 수 있다. 또한 Few-shot Learning 기반 접근을 적용하여 교량 뿐만 아니라 다양한 공종 내에서도 질의 생성을 가능하게 함으로써, 프로젝트의 확장성을 확보 가능하다.

키워드
BIM
지식그래프
GraphRAG
LPG
거대 언어 모델
Few-shot Learing
References
1

Beetz, J., Leeuwen, J.V., Vries, B.D. (2009) IfcOWL: A Case of Transforming EXPRESS Schemas into Ontologies, Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 23(1), pp.89~101.

10.1017/S0890060409000122
2

Brown, T.B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child, R., Ramesh, A., Ziegler, D.M., Wu, J., Winter, C., Hesse, C., Chen, M., Sigler, E., Litwin, M., Gray, S., Chess, B., Clark, J., Berner, C., McCandlish, S., Radford, A., Sutskever, I., Amodei, D. (2020) Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems (NeurIPS), 33. arXiv:2005.14165

3

Chen, X., Lu, J., Zhu, Y., Tang, Y. (2021) QAWizard: A Question Answering System for Building Information Models, Automation in Construction, 131, 103891.

4

Du, J., Li, X., Wang, B. (2024) Retrieval-Augmented Generation for AI-Generated Content: A Survey, arXiv preprint arXiv:2402.19473.

5

Edge, D., Trinh, H., Cheng, N., Bradley, J., Chao, A., Mody, A., Truitt, S., Metropolitansky, D., Ness, R.O., Larson, J. (2024). From Local to Global: A Graph RAG Approach to Query-Focused Summarization. arXiv preprint arXiv:2404.16130.

6

Francis, N., Green, A., Guagliardo, P., Libkin, L., Lindaaker, T., Marsault, V., Plantikow, S., Rydberg, M., Selmer, P., Taylor, A. (2018) Cypher: An Evolving Query Language for Property Graphs, Proceedings of the 2018 International Conference on Management of Data (SIGMOD'18), ACM, pp.1433~1445.

10.1145/3183713.3190657
7

Guo, D., Onstein, E., Rosa, A.D.L. (2020) An Approach of Automatic SPARQL Generation for BIM Data Extraction, Applied Sciences, 10(24), 8794.

10.3390/app10248794
8

Jeon, Y.H., Lee, S.H. (2025) A Comparative Study of Hybrid RAG and GraphRAG Approaches for Residential Defect Management, Korean Journal of Construction Engineering and Management, 26(1), pp.45~56.

9

Kang, T.W., Park, C.H. (2024) BIM Knowledge Expert Agent Based on Retrieval-Augmented Generation with Limited GPU Resources, Journal of Building Engineering, 84, 108512.

10

Lan, Y., He, G., Jiang, J., Jiang, J., Zhao, W.X., Wen, J.R. (2022). Complex knowledge base question answering: A survey. IEEE Transactions on Knowledge and Data Engineering, 35(11), pp.11196`11215.

10.1109/TKDE.2022.3223858
11

Larson, J., Truitt, S. (2024). GraphRAG: Unlocking LLM discovery on narrative private data. Microsoft Research Blog.

12

Mokhov, A. (2024) RAG-Based AI Assistant for Building Code Interpretation, Engineering Applications of Artificial Intelligence, 133, 108234.

13

OpenAI. (2024). GPT-4o mini: Advancing cost-efficient intelligence. https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/

14

Pauwels, P., Terkaj, W. (2016) EXPRESS to OWL for Construction Industry: Towards a Recommendable and Usable ifcOWL Ontology, Automation in Construction, 63, pp.100~133.

10.1016/j.autcon.2015.12.003
15

Peng, B., Zhu, Y., Liu, Y., Bo, X., Shi, H., Hong, C., Zhang, Y., Tang, S. (2024) Graph Retrieval-Augmented Generation: A Survey, arXiv preprint arXiv:2408.08921.

16

Shi, X., Wang, Y., Liu, J., Chen, Z. (2025) Knowledge Graph Enhanced Chain-of-Thought for Building Maintenance Document Generation, Automation in Construction, 159, 105198.

17

Unger, C., Buhmann, L., Lehmann, J., Ngonga Ngomo, A.C., Gerber, D., Cimiano, P. (2012) Template-based Question Answering over RDF Data, Proceedings of the 21st World Wide Web Conference 2012, WWW 2012, Lyon, France, ACM, pp.639~648.

10.1145/2187836.2187923
18

Wang, Z., Sacks, R., Ouyang, B., Ying, H., Borrmann, A. (2024). A framework for generic semantic enrichment of BIM models, Journal of Computing in Civil Engineering, 38(1), 04023038.

10.1061/JCCEE5.CPENG-5487
19

Xiong, Y., Li, M., Zhang, H., Wang, Q. (2025) DR-RAG: Digital Twin and RAG Integration for Aviation Design, Journal of Aerospace Engineering, 38(2), 04025015.

20

Yang, Z., Qi, P., Zhang, S., Bengio, Y., Cohen, W.W., Salakhutdinov, R., Manning, C.D. (2018). HotpotQA: A dataset for diverse, explainable multi-hop question answering. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing (pp.2369~2380).

10.18653/v1/D18-1259
21

Yih, W.T., Chang, M.W., He, X., Gao, J. (2015) Semantic Parsing via Staged Query Graph Generation: Question Answering with Knowledge Base, Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics, ACL 2015, Beijing, China, pp.1321~1331.

10.3115/v1/P15-1128
22

Yin, X., Gromann, D., Rudolph, S. (2023) GSP4BIM: Graph Neural Network Based Semantic Parsing for Building Information Models, Automation in Construction, 152, 104932.

10.1016/j.autcon.2023.104902
23

Yu, S.H., Kim, J.W., Park, H.S. (2025) Evaluation of GPT-4 Performance on Korean BIM Professional Examination with RAG Enhancement, Journal of Computing in Civil Engineering, 39(1), 04024042.

Information
  • Publisher :Computational Structural Engineering Institute of Korea
  • Publisher(Ko) :한국전산구조공학회
  • Journal Title :Journal of the Computational Structural Engineering Institute of Korea
  • Journal Title(Ko) :한국전산구조공학회 논문집
  • Volume : 39
  • No :1
  • Pages :59-70
  • Received Date : 2025-12-22
  • Revised Date : 2026-01-01
  • Accepted Date : 2026-01-06
  • DOI :https://doi.org/10.7734/COSEIK.2026.39.1.59
Journal Informaiton Journal of the Computational Structural Engineering Institute of Korea Journal of the Computational Structural Engineering Institute of Korea
Online Submission
  • NRF
  • KOFST
  • crossref crossmark
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close