SP
BravenNow
Qwen-BIM: developing large language model for BIM-based design with domain-specific benchmark and dataset
| USA | technology | ✓ Verified - arxiv.org

Qwen-BIM: developing large language model for BIM-based design with domain-specific benchmark and dataset

#Qwen-BIM #BIM #Large Language Models #Construction Industry #Digital Transformation #Domain-specific AI #Benchmark

📌 Key Takeaways

  • Qwen-BIM achieves a 21.0% average increase in G-Eval score compared to base LLM model
  • With only 14B parameters, Qwen-BIM performs comparably to general LLMs with 671B parameters for BIM tasks
  • The research introduces the first domain-specific LLM for BIM-based design
  • General LLMs are shown to be incompetent for domain-specific BIM tasks

📖 Full Retelling

Researchers Jia-Rui Lin, Yun-Hong Cai, Xiang-Rui Ni, Shaojie Zhou, and Peng Pan introduced Qwen-BIM, a specialized large language model for BIM-based design, in a paper published on arXiv on February 24, 2026, addressing the critical gap in domain-specific datasets and evaluation benchmarks that have hindered the performance of general language models in construction industry applications. As the construction industry advances toward digital transformation, Building Information Modeling (BIM) has emerged as a key driver supporting intelligent construction processes, though existing large language models have struggled to effectively handle domain-specific tasks without specialized training resources. The research team tackled this challenge by developing three critical components: an evaluation benchmark for BIM-based design with corresponding quantitative indicators, a method for generating textual data from BIM models to create specialized datasets, and a fine-tuning strategy specifically adapted for BIM-based design applications. Results from their comprehensive testing demonstrated that general large language models remain inadequate for specialized BIM tasks, highlighting the necessity of domain-specific approaches in AI applications for technical industries.

🏷️ Themes

Artificial Intelligence, Construction Technology, Digital Transformation

📚 Related People & Topics

Benchmark

Topics referred to by the same term

Benchmark may refer to:

View Profile → Wikipedia ↗
Building information modeling

Building information modeling

Process for digital management of built assets

Building information modeling (BIM) is an approach involving the generation and management of digital representations of the physical and functional characteristics of buildings or other physical assets and facilities. BIM is supported by various tools, processes, technologies and contracts. Buildi...

View Profile → Wikipedia ↗

Digital transformation

Adoption of digital technology by an organisation

Digital transformation (DT) is the process of adoption and implementation of digital technology by an organization in order to create new or modify existing products, services and operations by the means of translating business processes into a digital format. The goal for its implementation is to ...

View Profile → Wikipedia ↗
Construction

Construction

Process of building or assembling a building or infrastructure

Construction is the process involved in delivering buildings, infrastructure, industrial facilities, and associated activities through to the end of their life. It typically starts with planning, financing, and design that continues until the asset is built and ready for use. Construction also cover...

View Profile → Wikipedia ↗

Large language model

Type of machine learning model

A large language model (LLM) is a language model trained with self-supervised machine learning on a vast amount of text, designed for natural language processing tasks, especially language generation. The largest and most capable LLMs are generative pre-trained transformers (GPTs) that provide the c...

View Profile → Wikipedia ↗

Entity Intersection Graph

Connections for Benchmark:

🌐 Large language model 2 shared
🌐 Artificial intelligence 1 shared
🌐 Coordination failure 1 shared
🌐 Existential risk from artificial intelligence 1 shared
🌐 Game theory 1 shared
View full profile
Original Source
--> Computer Science > Artificial Intelligence arXiv:2602.20812 [Submitted on 24 Feb 2026] Title: Qwen-BIM: developing large language model for BIM-based design with domain-specific benchmark and dataset Authors: Jia-Rui Lin , Yun-Hong Cai , Xiang-Rui Ni , Shaojie Zhou , Peng Pan View a PDF of the paper titled Qwen-BIM: developing large language model for BIM-based design with domain-specific benchmark and dataset, by Jia-Rui Lin and 4 other authors View PDF Abstract: As the construction industry advances toward digital transformation, BIM (Building Information Modeling)-based design has become a key driver supporting intelligent construction. Despite Large Language Models have shown potential in promoting BIM-based design, the lack of specific datasets and LLM evaluation benchmarks has significantly hindered the performance of LLMs. Therefore, this paper addresses this gap by proposing: 1) an evaluation benchmark for BIM-based design together with corresponding quantitative indicators to evaluate the performance of LLMs, 2) a method for generating textual data from BIM and constructing corresponding BIM-derived datasets for LLM evaluation and fine-tuning, and 3) a fine-tuning strategy to adapt LLMs for BIM-based design. Results demonstrate that the proposed domain-specific benchmark effectively and comprehensively assesses LLM capabilities, highlighting that general LLMs are still incompetent for domain-specific tasks. Meanwhile, with the proposed benchmark and datasets, Qwen-BIM is developed and achieves a 21.0% average increase in G-Eval score compared to the base LLM model. Notably, with only 14B parameters, performance of Qwen-BIM is comparable to that of general LLMs with 671B parameters for BIM-based design tasks. Overall, this study develops the first domain-specific LLM for BIM-based design by introducing a comprehensive benchmark and high-quality dataset, which provide a solid foundation for developing BIM-related LLMs in various fields. Subjects: Artifici...
Read full article at source

Source

arxiv.org

More from USA

News from Other Countries

🇬🇧 United Kingdom

🇺🇦 Ukraine