Text mining and natural language processing in construction

Automation in Construction, Journal Year: 2023, Volume and Issue: 158, P. 105200 - 105200

Published: Nov. 21, 2023

Language: Английский

---

Artificial Intelligence Methods for the Construction and Management of Buildings

Sensors, Journal Year: 2023, Volume and Issue: 23(21), P. 8740 - 8740

Published: Oct. 26, 2023

Artificial intelligence covers a variety of methods and disciplines including vision, perception, speech dialogue, decision making planning, problem solving, robotics other applications in which self-learning is possible. The aim this work was to study the possibilities using AI algorithms at various stages construction ensure safety process. objects research were scientific publications about use artificial ways optimize To search for information, Scopus Web Science databases used period from early 1990s (the appearance first publication on topic) until end 2022. Generalization main method. It has been established that set technologies complement traditional human qualities, such as well analytical abilities. 3D modeling design buildings, machine learning conceptualization 3D, computer planning effective equipment, superintelligence have studied. proven automatic programming natural language processing, knowledge-based systems, robots, building maintenance, adaptive strategies, programming, genetic unmanned aircraft systems allow an evaluation construction. prospects are shown.

Language: Английский

---

Predicting Safety Accident Costs in Construction Projects Using Ensemble Data-Driven Models

Journal of Construction Engineering and Management, Journal Year: 2024, Volume and Issue: 150(7)

Published: April 16, 2024

The construction industry suffers from frequent and expensive safety accidents, significantly affecting project performance. Numerous data-driven classification models have been developed to categorize accident outcomes. While critical influencing factors provide insights for prevention, existing given less attention the cost of accidents—an important indicator management decisions. This study aims develop prediction that examine crucial precursors offering guidance prevention a financial perspective. collected 1,606 reports Chinese between 2005 2022 address this gap. Three ensemble methods, namely random forest, extreme gradient boosting regressor (XGBoost), natural (NGBoost) were employed models. Based on performance comparison, forest regression model was determined be best model. To extract attributes costs, utilized shapely additive explanations (SHAP) value analyze sensitivity influence input variables findings showed collapse has greatest impact as indicated by highest mean SHAP value, followed falling height. Furthermore, such year, supervision, drawing, plan are noteworthy in prediction. Safety department, protection, work conditions hold slightly higher degree compared contracting arrangement, culture, training examination, mechanical equipment output. provides dimension might overlooked investigation accidents provided will contribute development targeted strategies.

Language: Английский

---

Natural language processing-based ensemble technique to predict potential accident severity

International Journal of Systems Assurance Engineering and Management, Journal Year: 2025, Volume and Issue: unknown

Published: April 23, 2025

Language: Английский

---

An integration of intelligent approaches and economic criteria for predictive analytics of occupational accidents

Decision Analytics Journal, Journal Year: 2023, Volume and Issue: 9, P. 100357 - 100357

Published: Nov. 7, 2023

Occupational accidents are a significant concern, resulting in human suffering, economic crises, and social issues. Despite ongoing efforts to comprehend their causes predict occurrences, the use of machine learning models this domain remains limited. This study aims address gap by investigating intelligent approaches that incorporate criteria occupational accidents. Four algorithms, Random Forest (RF), Support Vector Machine (SVM), Multivariate Adaptive Regression Spline (MARS), M5 Tree Model (M5), were employed accidents, considering three criteria: basic income (BI), inflation index (II), price (PI). The focuses on identifying most suitable model for predicting frequency (FOA) determining with greatest influence. results reveal RF accurately predicts across all levels. Additionally, among criteria, II had impact findings suggest reduction FOA is unlikely coming years due increasing growth PI, coupled slight annual increase BI. Implementing appropriate countermeasures enhance workers' welfare, particularly low-income employees, crucial reducing research underscores potential preventing while highlighting critical role factors. It contributes valuable insights scholars, practitioners, policymakers develop effective strategies interventions improve workplace safety well-being.

Language: Английский

---

Applications of interpretable ensemble learning for workplace risk assessment: The Chinese coal industry as an example

Risk Analysis, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 21, 2025

Machine learning has demonstrated potential in addressing complex nonlinear changes risk assessment. However, further exploration is needed to enhance model interpretability and optimize performance. Therefore, this study aims develop a novel workplace assessment framework. By utilizing the SHapley Additive exPlanations (SHAP) analysis method ensemble algorithms, framework maps characteristic attributes levels. Reliability validation of critical attribute components are conducted using accidents Chinese coal enterprises as case study, which represents one most serious occupational hazards. The results indicate that issues algorithms yields capable accurately assessing understanding decision-making processes. Comparative experiments show achieves an accuracy up 98.3%, confirming its robust outcomes SHAP for feature importance facilitate identification explain causal relationships leading risk-level findings. This provides valuable accident prevention strategies minimize injuries losses.

Language: Английский

---

A review of machine learning for analysing accident reports in the construction industry

Journal of Information Technology in Construction, Journal Year: 2025, Volume and Issue: 30, P. 439 - 460

Published: April 1, 2025

Recently, there has been a growth in the research interest on applied machine learning (ML) safety analysis construction industry. The increased is part of search for improved prevention occupational accidents with focus text and natural language processing (NLP). However, ML-based approaches have less adapted compared to their perceived benefits due barriers implementation challenges analysing records sector. And current literature criticized lack clarity around description methodologies, interpretation, context application. Therefore, this work aims review latest developments applying ML accident report construction. A published reports was carried out organized terms data pre-processing, algorithms, testing further based structure. results found limitation related availability besides manual structuring use unsupervised reflect complexity handling textual data. Moreover, types happen proportionally varying frequencies need careful tackling outside basic assumptions pre-processing addition general comparative studies automated pipelines. also showed that mining (DM) were used especially semi-structured unstructured datasets reflecting maybe inefficient (NLP) application these learning. Among reviewed articles, only four six prototypes externally validated environment thus we propose future efforts would benefit from incorporating standardized development method explicit how recommendation informs decision making. Future should experiment ascertain different choices stage, validating performance models practices. Finally, are more advanced NLP methods could be if domain specific repositories available such as relation extraction various advances explored including large (LLMs).

Language: Английский

---

Transforming workplace safety through leveraging predictive analytics and explainable AI in steel industries

Proceedings of the Institution of Mechanical Engineers Part O Journal of Risk and Reliability, Journal Year: 2025, Volume and Issue: unknown

Published: April 17, 2025

Workers in the steel manufacturing plants often confront perilous working conditions characterized by limited visibility, potential hazards from heavy machinery interacting with pedestrian staff, and dangerous dynamicity of processes. Such environments involve repetitive tasks, extreme temperatures, high noise levels, challenging surroundings, fostering situational behavioral risks that escalate likelihood accidents leading to injuries, illnesses, or fatalities. Therefore, it is imperative scrutinize safety incidents within industries mitigate enhance measures proactively. This study employs Machine Learning (ML) develop predictive models using a dataset comprising 3600 workplace reported year 2018 2022 three integrated India. The aim identify indicative unsafe events situations based on different ML models. Five were compared viz. Random Forest, Gradient Boosting, Support Vector Machine, Decision Tree, K-Nearest Neighbor. Forest emerged as most effective, achieving 86.52% accuracy 100% AUC score three-class classification. classification accident types provides valuable insights into risks, enabling proactive prevent future incidents. Through appropriate identification lead specific accidents, this research offers data-driven approach protocols. Furthermore, contributes significantly Explainable AI (XAI), such Local Interpretable Model-Agnostic Explanations (LIME), particularly enhancing approaches Indian industry.

Language: Английский

---

A hybrid approach integrating case mining (CM) and the Copula Bayesian Network (CBN) for accident causation probabilistic reasoning of building construction collapses

Reliability Engineering & System Safety, Journal Year: 2024, Volume and Issue: 252, P. 110469 - 110469

Published: Aug. 26, 2024

Language: Английский

---

Research on the Prediction of Sustainable Safety Production in Building Construction Based on Text Data

Sustainability, Journal Year: 2024, Volume and Issue: 16(12), P. 5081 - 5081

Published: June 14, 2024

Given the complexity and variability of modern construction projects, safety risk management has become increasingly challenging, while traditional methods exhibit deficiencies in handling complex dynamic environments, particularly those involving unstructured text data. Consequently, this study proposes a data-based prediction method for building safety. Initially, heuristic Chinese automatic word segmentation, which incorporates mutual information, information entropy statistics, TF-IDF algorithm, preprocesses data to extract factor keywords construct accident attribute variables. At same time, Spearman correlation coefficient is utilized eliminate multicollinearity between feature Next, XGBoost algorithm employed develop model predicting risks associated with safe production. Its performance optimized through three experimental scenarios. The results indicate that achieves satisfactory overall after hyperparameter tuning, accuracy F1 score reaching approximately 86%. Finally, SHAP interpretation technique identifies critical factors influencing production construction, highlighting project managers’ attention safety, government regulation, design, emergency response as determinants severity. main objective minimize human intervention assessment using rich empirical knowledge embedded text, aim reducing accidents promoting sustainable development industry. This not only enables paradigm shift toward intelligent control but also provides theoretical practical insights into decision-making technical support

Language: Английский

---