Zobrazit v EDS

Improving the Predictability of Hazardous Scenarios by Natural Language Processing: The case of accidents during lifting operations on ships and offshore platforms

Uloženo v:
Podrobná bibliografie
Název: Improving the Predictability of Hazardous Scenarios by Natural Language Processing: The case of accidents during lifting operations on ships and offshore platforms
Autoři: Cárdenas, Ibsen Chivatá, Kozin, Igor
Informace o vydavateli: Research Publishing Services, Singapore, 2025.
Rok vydání: 2025
Témata: Hazard identification, Predicitive accident analysis, Natural Language Processing
Popis: The completeness and high predictability of hazardous scenarios by hazard identification methods are issues in risk analyses. A way to the improvement is to carry out both an exhaustive - to the extent possible - post-accident and predictive accident analysis. Currently, Natural Language Processing (NLP) allows quick processing of many accident reports. In combination with graphical tools, it is now even possible to automatically output causal diagrammatic models of accidents and visualize them on a multi-scenario accident diagram. A step forward is the application of NLP to support predictive analysis. Predictive accident analysis focuses on identifying deviations from expected or normal conditions, the subsequent events following these deviations, and their interactions leading to an accident. The expected or normal conditions are typically outlined in specifications and procedures. This paper demonstrates how NLP can assist hazard identification and predictive accident analysis during lifting operations on ships and offshore platforms.
Druh dokumentu: Conference object
Jazyk: English
DOI: 10.3850/978-981-94-3281-3_esrel-sra-e2025-p5041-cd
Přístupová URL adresa: https://portal.findresearcher.sdu.dk/da/publications/b6762e5a-61c8-43d4-8e96-5ba7f93b4f62
Přístupové číslo: edsair.od......3062..d5ca3840ecce608dcfba7a0e668428a3
Databáze: OpenAIRE
Popis
Abstrakt:The completeness and high predictability of hazardous scenarios by hazard identification methods are issues in risk analyses. A way to the improvement is to carry out both an exhaustive - to the extent possible - post-accident and predictive accident analysis. Currently, Natural Language Processing (NLP) allows quick processing of many accident reports. In combination with graphical tools, it is now even possible to automatically output causal diagrammatic models of accidents and visualize them on a multi-scenario accident diagram. A step forward is the application of NLP to support predictive analysis. Predictive accident analysis focuses on identifying deviations from expected or normal conditions, the subsequent events following these deviations, and their interactions leading to an accident. The expected or normal conditions are typically outlined in specifications and procedures. This paper demonstrates how NLP can assist hazard identification and predictive accident analysis during lifting operations on ships and offshore platforms.
DOI:10.3850/978-981-94-3281-3_esrel-sra-e2025-p5041-cd