In this work, we propose a method to identify emerging safety threats in the NASA’s aviation safety reporting system using topic modeling. The proposed method serves both to connect different topics over time, but also to facilitate quick exploration and navigation. We use COVID-19 as a use case to identify if emerging safety threats can be identified over time. We found not only the topics’ top terms indicated the prevalence of COVID-19 reports, but also they accounted for 71% of manually curated reports, offering an automated means to identify emerging threats.
Identifying Emerging Safety Threats through Topic Modeling in the Aviation Safety Reporting System: A COVID-19 Study
2021-10-03
1398392 byte
Conference paper
Electronic Resource
English