Zezhong Wang

The transition toward Industry 5.0 requires safety management systems that are not only intelligent and data-driven but also human-centric, resilient, and aligned with sustainable operations. However, conventional safety risk assessment in steel manufacturing remains heavily dependent on expert judgment and often lacks the adaptability required to address complex and dynamic production environments. This study proposes a text-driven intelligent framework for sustainable safety risk governance by integrating natural language processing, topic modeling, objective indicator weighting, and Bayesian decision fusion. The framework establishes a closed-loop process encompassing risk identification, quantitative assessment, risk classification, and hierarchical control. It automatically extracts risk-related information from unstructured safety records, maps the identified hazards onto a human–machine–environment–management structure, quantifies multidimensional risk indicators, and translates assessment outcomes into differentiated control measures. The framework was evaluated using field safety records collected from Tianjin Iron and Steel Group. The topic modeling results identified four major dimensions of operational risk, while the CRITIC–Bayesian weighting mechanism combined data-driven indicator differentiation with context-sensitive probabilistic reasoning. Following its integration into the company’s intelligent safety management platform and one year of operational use, the framework reduced the time required to formulate safety inspection plans by 70%, supported dynamic four-level risk classification, and achieved a 97% task completion rate. The number of recorded safety accidents also decreased by 35% compared with the pre-deployment baseline. These findings demonstrate that unstructured safety text can be transformed into actionable risk intelligence, enhancing the proactive, systematic, and adaptive governance of safety risks. The proposed framework provides a practical pathway for advancing human-centric safety management, operational resilience, and sustainable production in the steel industry.

Based on a spatial clustering and partitioned stacking ensemble model, this study addresses the limitations of traditional geoweighting regression in capturing nonlinear location premiums and submarket heterogeneity within urban real estate markets. It proposes a two - stage modeling framework: “spatial clustering → partitioned differentiated stacking ensemble.” Using long - term multi - source transaction data for Beijing's secondary housing market, the study divides the market into three spatially heterogeneous submarkets: core, near - suburban, and far - suburban. Stacked ensemble models based on ElasticNet, XGBoost, LightGBM, and Random Forest are constructed within each submarket.Factor analysis extracts interpretable common factors, which are combined with Lasso and SHAP for feature selection and impact mechanism analysis.Results indicate that the zoned stacking model performs exceptionally well across all three submarkets, achieving an R² of 0.916 in the core urban area. Significant nonlinear location premiums exist within the core urban area.The multi - level interpretability framework reveals the differentiated effects of location and scale factors across different submarkets.This study advances from “global modeling” to “spatial zoning + adaptive ensemble,” providing a viable tool for refined valuation and risk management in highly heterogeneous markets.