A computer leaning approach to obtain safety information from multi-lingual accident reports

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Accident reports provide a valuable source of data for any safety management system. In multi-lingual jurisdictions, accident reports can be provided in more than one language. For example the Swiss transport authority collects accident reports that are written in either German, French, or Italian. The unstructured nature of free-text makes it difficult to extract information from large numbers of accident reports. Machine-reading of text is an emerging area of research, however there are few instances of information being extracted from text in more than one language.

This paper introduces an ontology-based interactive learning method between a human and computer software to identify safety-related information by analysing text written in three different languages. The results of the method were analysed by fluent speakers of each language, who rated the overall accuracy of the method to be 98.5%.

The method stores and processes the data in a NoSQL graph database, which provides a powerful tool to readily integrate the analysis with other data sources, for example train movement data, passenger census data, or even comparative data from other railways.
LanguageEnglish
Title of host publicationSafety and Reliability – Safe Societies in a Changing World
Subtitle of host publicationProceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway
PublisherCRC Press / Balkema, Taylor and Francis group
Chapter389
Pages3107-3114
Number of pages8
ISBN (Print)9780815386827
Publication statusPublished - 18 Jun 2018
EventAnnual European Safety and Reliability Conference - Norwegian University of Science and Technology, Trodheim, Norway
Duration: 17 Jun 201821 Jun 2018
https://www.ntnu.edu/web/esrel2018/home (Link to Conference Website )

Conference

ConferenceAnnual European Safety and Reliability Conference
Abbreviated titleESREL
CountryNorway
CityTrodheim
Period17/06/1821/06/18
Internet address

Fingerprint

Accidents
Ontology

Cite this

Hughes, P., Figueres Esteban, M., El Rashidy, R., Van Gulijk, C., & Slovak, R. (2018). A computer leaning approach to obtain safety information from multi-lingual accident reports. In Safety and Reliability – Safe Societies in a Changing World: Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway (pp. 3107-3114). CRC Press / Balkema, Taylor and Francis group.
Hughes, Peter ; Figueres Esteban, Miguel ; El Rashidy, Rawia ; Van Gulijk, Coen ; Slovak, R. / A computer leaning approach to obtain safety information from multi-lingual accident reports. Safety and Reliability – Safe Societies in a Changing World: Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway. CRC Press / Balkema, Taylor and Francis group, 2018. pp. 3107-3114
@inproceedings{5cb0e418a5a4458e85979f952f61e4eb,
title = "A computer leaning approach to obtain safety information from multi-lingual accident reports",
abstract = "Accident reports provide a valuable source of data for any safety management system. In multi-lingual jurisdictions, accident reports can be provided in more than one language. For example the Swiss transport authority collects accident reports that are written in either German, French, or Italian. The unstructured nature of free-text makes it difficult to extract information from large numbers of accident reports. Machine-reading of text is an emerging area of research, however there are few instances of information being extracted from text in more than one language.This paper introduces an ontology-based interactive learning method between a human and computer software to identify safety-related information by analysing text written in three different languages. The results of the method were analysed by fluent speakers of each language, who rated the overall accuracy of the method to be 98.5{\%}.The method stores and processes the data in a NoSQL graph database, which provides a powerful tool to readily integrate the analysis with other data sources, for example train movement data, passenger census data, or even comparative data from other railways.",
keywords = "Informatics",
author = "Peter Hughes and {Figueres Esteban}, Miguel and {El Rashidy}, Rawia and {Van Gulijk}, Coen and R. Slovak",
year = "2018",
month = "6",
day = "18",
language = "English",
isbn = "9780815386827",
pages = "3107--3114",
booktitle = "Safety and Reliability – Safe Societies in a Changing World",
publisher = "CRC Press / Balkema, Taylor and Francis group",

}

Hughes, P, Figueres Esteban, M, El Rashidy, R, Van Gulijk, C & Slovak, R 2018, A computer leaning approach to obtain safety information from multi-lingual accident reports. in Safety and Reliability – Safe Societies in a Changing World: Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway. CRC Press / Balkema, Taylor and Francis group, pp. 3107-3114, Annual European Safety and Reliability Conference, Trodheim, Norway, 17/06/18.

A computer leaning approach to obtain safety information from multi-lingual accident reports. / Hughes, Peter; Figueres Esteban, Miguel; El Rashidy, Rawia; Van Gulijk, Coen; Slovak, R.

Safety and Reliability – Safe Societies in a Changing World: Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway. CRC Press / Balkema, Taylor and Francis group, 2018. p. 3107-3114.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A computer leaning approach to obtain safety information from multi-lingual accident reports

AU - Hughes,Peter

AU - Figueres Esteban,Miguel

AU - El Rashidy,Rawia

AU - Van Gulijk,Coen

AU - Slovak,R.

PY - 2018/6/18

Y1 - 2018/6/18

N2 - Accident reports provide a valuable source of data for any safety management system. In multi-lingual jurisdictions, accident reports can be provided in more than one language. For example the Swiss transport authority collects accident reports that are written in either German, French, or Italian. The unstructured nature of free-text makes it difficult to extract information from large numbers of accident reports. Machine-reading of text is an emerging area of research, however there are few instances of information being extracted from text in more than one language.This paper introduces an ontology-based interactive learning method between a human and computer software to identify safety-related information by analysing text written in three different languages. The results of the method were analysed by fluent speakers of each language, who rated the overall accuracy of the method to be 98.5%.The method stores and processes the data in a NoSQL graph database, which provides a powerful tool to readily integrate the analysis with other data sources, for example train movement data, passenger census data, or even comparative data from other railways.

AB - Accident reports provide a valuable source of data for any safety management system. In multi-lingual jurisdictions, accident reports can be provided in more than one language. For example the Swiss transport authority collects accident reports that are written in either German, French, or Italian. The unstructured nature of free-text makes it difficult to extract information from large numbers of accident reports. Machine-reading of text is an emerging area of research, however there are few instances of information being extracted from text in more than one language.This paper introduces an ontology-based interactive learning method between a human and computer software to identify safety-related information by analysing text written in three different languages. The results of the method were analysed by fluent speakers of each language, who rated the overall accuracy of the method to be 98.5%.The method stores and processes the data in a NoSQL graph database, which provides a powerful tool to readily integrate the analysis with other data sources, for example train movement data, passenger census data, or even comparative data from other railways.

KW - Informatics

M3 - Conference contribution

SN - 9780815386827

SP - 3107

EP - 3114

BT - Safety and Reliability – Safe Societies in a Changing World

PB - CRC Press / Balkema, Taylor and Francis group

ER -

Hughes P, Figueres Esteban M, El Rashidy R, Van Gulijk C, Slovak R. A computer leaning approach to obtain safety information from multi-lingual accident reports. In Safety and Reliability – Safe Societies in a Changing World: Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway. CRC Press / Balkema, Taylor and Francis group. 2018. p. 3107-3114