Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives

Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives

Three methods are demonstrated for automated classification of aviation safety narratives within an existing complex taxonomy. Utilizing latent semantic analysis trained against 4497 narratives at the sentence level, primary problem and contributing factor labels were assessed. Results from a sample...

Full description

Bibliographic Details
Main Author: Saul D. Robinson
Format: Article
Language:English
Published: MDPI AG 2018-07-01
Series:Safety
Subjects:
latent semantic analysis (LSA)
adaptive taxonomy
safety
automatic indexing
machine learning
Online Access:http://www.mdpi.com/2313-576X/4/3/30
id doaj-fc2950898cc84da9b4bd5d6a027b1f49
record_format Article
spelling doaj-fc2950898cc84da9b4bd5d6a027b1f492020-11-24T21:36:19ZengMDPI AGSafety2313-576X2018-07-01433010.3390/safety4030030safety4030030Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety NarrativesSaul D. Robinson0Parks College of Engineering, Aviation and Technology, Saint Louis University, Saint Louis, MO 63103, USAThree methods are demonstrated for automated classification of aviation safety narratives within an existing complex taxonomy. Utilizing latent semantic analysis trained against 4497 narratives at the sentence level, primary problem and contributing factor labels were assessed. Results from a sample of 2987 narratives provided a mean unsupervised categorization precision of 0.35% and recall of 0.78% for contributing-factors within the taxonomy. Categorization of the primary problem at the sentence level resulted in a modal accuracy of 0.46%. Overall, the results suggested that the demonstrated approaches were viable in bringing additional tools and insights to safety researchers.http://www.mdpi.com/2313-576X/4/3/30latent semantic analysis (LSA)adaptive taxonomysafetyautomatic indexingmachine learning
collection DOAJ
language English
format Article
sources DOAJ
author Saul D. Robinson
spellingShingle Saul D. Robinson
Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
Safety
latent semantic analysis (LSA)
adaptive taxonomy
safety
automatic indexing
machine learning
author_facet Saul D. Robinson
author_sort Saul D. Robinson
title Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
title_short Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
title_full Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
title_fullStr Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
title_full_unstemmed Multi-Label Classification of Contributing Causal Factors in Self-Reported Safety Narratives
title_sort multi-label classification of contributing causal factors in self-reported safety narratives
publisher MDPI AG
series Safety
issn 2313-576X
publishDate 2018-07-01
description Three methods are demonstrated for automated classification of aviation safety narratives within an existing complex taxonomy. Utilizing latent semantic analysis trained against 4497 narratives at the sentence level, primary problem and contributing factor labels were assessed. Results from a sample of 2987 narratives provided a mean unsupervised categorization precision of 0.35% and recall of 0.78% for contributing-factors within the taxonomy. Categorization of the primary problem at the sentence level resulted in a modal accuracy of 0.46%. Overall, the results suggested that the demonstrated approaches were viable in bringing additional tools and insights to safety researchers.
topic latent semantic analysis (LSA)
adaptive taxonomy
safety
automatic indexing
machine learning
url http://www.mdpi.com/2313-576X/4/3/30
work_keys_str_mv AT sauldrobinson multilabelclassificationofcontributingcausalfactorsinselfreportedsafetynarratives
_version_ 1725941694629675008

Similar Items