July Bias Macedo
Development of Natural Language Processing-based Solutions for Risk Analysis: application to a hydropower company and an O&G industry (2022)
Supervisors: Prof. Dr. Márcio José das Chagas Moura, Prof. Dr. Enrico Zio
Keywords: Risk analysis; accident investigation reports; natural language processing; text mining; oil refineries; hydroelectric power company
This research aims to address the challenges in Risk Analysis (RA) by developing innovative Natural Language Processing (NLP) solutions. To improve RA, the study develops Natural Language Processing (NLP) models to extract and organize information from textual data, including proactive data from preliminary risk studies and accident investigation reports. Two methodologies are presented: one identifies issues in accident reports, aiding safety technicians in proposing improvements, while the other automates risk feature identification for the initial stage of Quantitative Risk Analysis (QRA) in Oil and Gas industries. The research aims to prevent occupational and major accidents, reducing environmental, economic, and human losses.
More information on July's work
References and links:
- July Bias Macêdo, Márcio das Chagas Moura, Diego Aichele, Isis Didier Lins, Identification of risk features using text mining and BERT-based models: Application to an oil refinery, Process Safety and Environmental Protection, Volume 158, 2022, Pages 382-399, ISSN 0957-5820, https://doi.org/10.1016/j.psep.2021.12.025.
- July B. Macedo, Plinio M. S. Ramos, Caio B. S. Maior, Márcio J. C. Moura, Isis D. Lins & Romulo F. T. Vilela (2023) Identifying low-quality patterns in accident reports from textual data, International Journal of Occupational Safety and Ergonomics, 29:3, 1088-1100, DOI: 10.1080/10803548.2022.2111847
- Ramos PM, Macedo JB, Maior CB, Moura MC, Lins ID. Combining BERT with numerical variables to classify injury leave based on accident description. Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability. 2022;0(0). doi:10.1177/1748006X221140194