Guozong Zhang

In the construction sector, technical bid evaluation is crucial for maintaining equity, evaluating contractor capabilities, and allocating resources efficiently. However, traditional methods rely heavily on manual expert review, which can introduce subjective biases and efficiency bottlenecks. While the existing literature predominantly emphasizes multi-criteria decision-making (MCDM) techniques, automated content-level scoring mechanisms remain largely underexplored. This study aims to develop an evidence-grounded automated technical bid evaluation framework by integrating Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs), addressing the challenges of criterion–evidence alignment and traceable scoring in engineering document evaluation. We propose a criterion-guided framework that integrates a structured scoring criteria database, semantic retrieval, and LLM-based reasoning to identify relevant evidence and generate evidence-supported evaluation results. Empirical results on 37 real-world technical bids indicate that the proposed framework achieves better performance over baseline models in terms of absolute error (MAE), concordance (Lin’s CCC), and citation accuracy (CA), while maintaining comparable ranking consistency. The contributions of this study lie in adapting RAG and LLM techniques to the engineering technical bid evaluation scenario, establishing a criterion-guided and evidence-grounded evaluation mechanism to enhance scoring traceability, and providing empirical validation using real-world engineering bid documents. This study provides empirical evidence for automated bid scoring, offering a practical technical reference for the digital transformation of construction bid evaluation.

Frequent accidents in the construction sector arise from the dynamic coupling of multiple risk factors, while conventional single-factor approaches fail to capture the underlying complexity. Drawing on 702 accident investigation reports, this study develops an intelligent, data-driven framework that integrates large language model–based risk identification with association rule mining to systematically uncover risk factors and their coupling patterns. A DeepSeek-based model is employed to extract risk factors from unstructured text, followed by cosine similarity–based optimization to refine factor representations. The FP-Growth algorithm is then applied to identify strong association rules among risk factors. The results reveal that deficiencies in the management dimension account for 68.30% of all identified risks, with inadequate safety education and training emerging as the central hub in the risk coupling network, which is further corroborated by complex network analysis. Moreover, a cascading transmission pathway is identified, whereby environmental deficiencies induce weakened safety awareness, which in turn leads to unsafe behaviors. These findings further demonstrate the nonlinear amplification effects arising from concurrent management failures. By establishing a transformation pathway from unstructured textual data to structured risk knowledge, this study provides a robust, data-driven foundation for precise risk identification and systematic prevention in construction safety management.

Urban regeneration has increasingly become a global strategy for promoting sustainable urban development, with the renewal of deteriorating residential communities serving as a key dimension of this process. Within the framework of a people-centered development paradigm, growing attention has been directed toward the necessity of securing residents’ satisfaction in community renewal initiatives. This study employs advanced textual analysis of resident submissions collected from government–citizen interaction platforms to investigate the determinants of satisfaction with renewal projects. Leveraging the semantic comprehension capabilities of large language models (LLMs), we identify both salient keywords and sentiment orientations embedded in residents’ narratives. Guided by the theoretical framework of resident satisfaction, the extracted keywords are organized into seven thematic domains: basic infrastructure improvement, quality-enhancement renovation, solicitation of residents’ preferences, residents’ decision-making power, policy transparency, construction governance, and community-level communication. Regression modeling is subsequently applied to assess the relative influence of these thematic domains on residents’ satisfaction. The findings suggest that insufficient integration of residents’ preferences at the preliminary stages of participation constitutes a principal source of dissatisfaction during the implementation of renewal projects. Furthermore, the study compares Latent Dirichlet Allocation (LDA) topic modeling with LLMs-based topic clustering, revealing the latter’s superior capacity to capture thematic structures in complex, long-form textual data. These results underscore the potential of LLMs to enhance the analytical rigor of research on urban regeneration and citizen participation.

In the context of China’s rapid urbanization and the era of stock planning, urban renewal policies play a significant role in enhancing urban quality and promoting sustainable development. To reveal the thematic structure and evolution of China’s urban renewal policy system, this study applies the BERTopic model to conduct semantic mining and evolutionary analysis on 1144 policy documents issued by central and local governments. Research findings: The study identifies 34 distinct themes in urban renewal policies, grouped into five main directions: Spatial Improvement and Facility Upgrades, Project Collaboration and Approval, Land Acquisition and Compensation, Fiscal Incentives and Funding Support, and Institutional Guarantees and Governance. Each of these directions exhibits distinct evolutionary trends over time. While urban renewal policies in the Central, Western, Eastern, and Northeastern regions share common characteristics in key aspects such as land acquisition and compensation, funding assurance, and residential quality enhancement, they also reflect regional differences due to varying stages of development, economic conditions, and geographic factors. This demonstrates both the shared and distinct policy focus areas across different regions of China. By identifying underlying themes and their trajectories, this study provides critical insights into the structural characteristics of urban renewal policies and offers valuable references for government authorities to improve and optimize policy systems. At the same time, it provides the Chinese experience for urban renewal in other countries.

Prefabricated construction represents an advanced approach to sustainable development, and safety issues in prefabricated construction projects have drawn widespread attention. Safety accident case reports contain a wealth of safety knowledge, and extracting and learning from such historical reports can significantly enhance safety management capabilities. However, these texts are often semantically complex and lengthy, posing challenges for traditional Information Extraction (IE) methods. This study focuses on the challenge of Named Entity Recognition (NER) in long texts under complex engineering contexts and proposes a novel model that integrates Modern Bidirectional Encoder Representations from Transformers (ModernBERT),Bidirectional Long Short-Term Memory (BiLSTM), andConditional Random Field (CRF). A comparative analysis with current mainstream methods is conducted. The results show that the proposed model achieves an F1 score of 0.6234, outperforming mainstream baseline methods. Notably, it attains F1 scores of 0.95 and 0.92 for the critical entity categories “Consequence” and “Type,” respectively. The model maintains stable performance even under semantic noise interference, demonstrating strong robustness in processing unstructured and highly heterogeneous engineering texts. Compared with existing long-text NER models, the proposed method exhibits superior semantic parsing ability in engineering contexts. This study enhances information extraction methods and provides solid technical support for constructing safety knowledge graphs in prefabricated construction, thereby advancing the level of intelligence in the construction industry.

As projects in the architecture, engineering, construction, and operations (AECO) industry grow in complexity and scale, there is an urgent need for more effective information management and intelligent decision-making. This study investigates the potential of large language models (LLMs) to address these challenges by systematically reviewing their core technologies, application scenarios, and integration approaches in AECO. Using a literature-based review methodology, this paper examines how LLMs—built on Transformer architecture and powered by deep learning and natural language processing—can process complex unstructured data and support a wide range of tasks, including contract analysis, construction scheduling, risk assessment, and operations and maintenance. This study finds that while LLMs offer substantial promise for enhancing productivity and automation in AECO workflows, several obstacles remain, such as data quality issues, computational demands, limited adaptability, integration barriers, and ethical concerns. The paper concludes that future research should focus on improving model efficiency, enabling multimodal data fusion, and enhancing compatibility with existing industry tools to realize the full potential of LLMs and support the digital transformation of the AECO sector.