Natural Language Processing (NLP) tools for the analysis of incident and accident reports - HAL-SHS - Sciences de l'Homme et de la Société Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Natural Language Processing (NLP) tools for the analysis of incident and accident reports

Résumé

The human factor field is expected to evolve due to the development of Natural Language Processing tools which allow for new approaches to handle natural language data. In the current project, we use NLP methods to facilitate experience feedback in the field of civil aviation safety. In this paper, we present how NLP methods based on the extraction of textual information from the Air France ASR can contribute to (i) the improvement of the reliability of the coding, facilitating the coding itself, (ii) the analysis of reports regardless of the categorization in order to expand the analysis perimeter and to avoid the inherent limitations of the codification.
Fichier principal
Vignette du fichier
HCI-Aero12_CFH_v2_VFIN.pdf (322.5 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

halshs-00953658 , version 1 (28-02-2014)

Identifiants

  • HAL Id : halshs-00953658 , version 1

Citer

Christophe Pimm, Céline Raynal, Nikola Tulechki, Eric Hermann, Grégory Caudy, et al.. Natural Language Processing (NLP) tools for the analysis of incident and accident reports. International Conference on Human-Computer Interaction in Aerospace (HCI-Aero), 2012, Brussels, Belgium. ⟨halshs-00953658⟩
941 Consultations
705 Téléchargements

Partager

Gmail Facebook X LinkedIn More