Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows

Julián Zapata; Andreas Søeborg Kirkedal

Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows

Julián Zapata, Andreas Søeborg Kirkedal

University of Ottawa

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

140 Downloads (Pure)

Abstract

In this paper, we report on a two-part experiment aiming to assess and compare the performance of two types of automatic speech recognition (ASR) systems on two different computational platforms when used to augment dictation workflows. The experiment was performed with a sample of speakers of three major languages and with different linguistic profiles: non-native English speakers; non-native French speakers; and native Spanish speakers. The main objective of this experiment is to examine ASR performance in translation dictation (TD) and medical dictation (MD) workflows without manual transcription vs. with transcription. We discuss the advantages and drawbacks of a particular ASR approach in different computational platforms when used by various speakers of a given language, who may have different accents and levels of proficiency in that language, and who may have different levels of competence and experience dictating large volumes of text, and with ASR technology. Lastly, we enumerate several areas for future research.

Original language	English
Title of host publication	Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015
Editors	Beáta Megyesi
Number of pages	10
Place of Publication	Linköping
Publisher	Linköping University Electronic Press
Publication date	2015
Pages	201-210
ISBN (Print)	9789175190983
ISBN (Electronic)	9789175190983
Publication status	Published - 2015
Event	NODALIDA 2015: The 20th Nordic Conference of Computational Linguistics - Vilnius, Lithuania Duration: 11 May 2015 → 13 May 2015 Conference number: 20 http://www2.lingfil.uu.se/nodalida2015/

Conference

Conference	NODALIDA 2015: The 20th Nordic Conference of Computational Linguistics
Number	20
Country/Territory	Lithuania
City	Vilnius
Period	11/05/2015 → 13/05/2015
Internet address	http://www2.lingfil.uu.se/nodalida2015/

Series	NEALT (Northern European Association of Language Technology) Proceedings Series
Volume	23
ISSN	1736-6305

Access to Document

andreas_søeborg_kirkedal_assessing_the_performance_publishersversionFinal published version, 974 KBLicence: Unspecified

http://www.ep.liu.se/ecp/109/025/ecp15109025.pdfLicence: Unspecified

Handle.net

Persistent link

Cite this

Zapata, J., & Kirkedal, A. S. (2015). Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows. In B. Megyesi (Ed.), Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015 (pp. 201-210). Linköping University Electronic Press. http://www.ep.liu.se/ecp/109/025/ecp15109025.pdf

Zapata, Julián ; Kirkedal, Andreas Søeborg. / Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows. Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015. editor / Beáta Megyesi. Linköping : Linköping University Electronic Press, 2015. pp. 201-210 (NEALT (Northern European Association of Language Technology) Proceedings Series, Vol. 23).

@inproceedings{5343211789794d91aaebabd9bf490082,

title = "Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows",

abstract = "In this paper, we report on a two-part experiment aiming to assess and compare the performance of two types of automatic speech recognition (ASR) systems on two different computational platforms when used to augment dictation workflows. The experiment was performed with a sample of speakers of three major languages and with different linguistic profiles: non-native English speakers; non-native French speakers; and native Spanish speakers. The main objective of this experiment is to examine ASR performance in translation dictation (TD) and medical dictation (MD) workflows without manual transcription vs. with transcription. We discuss the advantages and drawbacks of a particular ASR approach in different computational platforms when used by various speakers of a given language, who may have different accents and levels of proficiency in that language, and who may have different levels of competence and experience dictating large volumes of text, and with ASR technology. Lastly, we enumerate several areas for future research.",

author = "Juli{\'a}n Zapata and Kirkedal, {Andreas S{\o}eborg}",

year = "2015",

language = "English",

isbn = "9789175190983",

series = "NEALT (Northern European Association of Language Technology) Proceedings Series",

pages = "201--210",

editor = "Be{\'a}ta Megyesi",

booktitle = "Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015",

publisher = "Link{\"o}ping University Electronic Press",

address = "Sweden",

note = "NODALIDA 2015: The 20th Nordic Conference of Computational Linguistics ; Conference date: 11-05-2015 Through 13-05-2015",

url = "http://www2.lingfil.uu.se/nodalida2015/",

}

Zapata, J & Kirkedal, AS 2015, Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows. in B Megyesi (ed.), Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015. Linköping University Electronic Press, Linköping, NEALT (Northern European Association of Language Technology) Proceedings Series, vol. 23, pp. 201-210, NODALIDA 2015: The 20th Nordic Conference of Computational Linguistics, Vilnius, Lithuania, 11/05/2015. <http://www.ep.liu.se/ecp/109/025/ecp15109025.pdf>

Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows. / Zapata, Julián; Kirkedal, Andreas Søeborg.
Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015. ed. / Beáta Megyesi. Linköping: Linköping University Electronic Press, 2015. p. 201-210 (NEALT (Northern European Association of Language Technology) Proceedings Series, Vol. 23).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows

AU - Zapata, Julián

AU - Kirkedal, Andreas Søeborg

N1 - Conference code: 20

PY - 2015

Y1 - 2015

N2 - In this paper, we report on a two-part experiment aiming to assess and compare the performance of two types of automatic speech recognition (ASR) systems on two different computational platforms when used to augment dictation workflows. The experiment was performed with a sample of speakers of three major languages and with different linguistic profiles: non-native English speakers; non-native French speakers; and native Spanish speakers. The main objective of this experiment is to examine ASR performance in translation dictation (TD) and medical dictation (MD) workflows without manual transcription vs. with transcription. We discuss the advantages and drawbacks of a particular ASR approach in different computational platforms when used by various speakers of a given language, who may have different accents and levels of proficiency in that language, and who may have different levels of competence and experience dictating large volumes of text, and with ASR technology. Lastly, we enumerate several areas for future research.

AB - In this paper, we report on a two-part experiment aiming to assess and compare the performance of two types of automatic speech recognition (ASR) systems on two different computational platforms when used to augment dictation workflows. The experiment was performed with a sample of speakers of three major languages and with different linguistic profiles: non-native English speakers; non-native French speakers; and native Spanish speakers. The main objective of this experiment is to examine ASR performance in translation dictation (TD) and medical dictation (MD) workflows without manual transcription vs. with transcription. We discuss the advantages and drawbacks of a particular ASR approach in different computational platforms when used by various speakers of a given language, who may have different accents and levels of proficiency in that language, and who may have different levels of competence and experience dictating large volumes of text, and with ASR technology. Lastly, we enumerate several areas for future research.

M3 - Article in proceedings

SN - 9789175190983

T3 - NEALT (Northern European Association of Language Technology) Proceedings Series

SP - 201

EP - 210

BT - Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015

A2 - Megyesi, Beáta

PB - Linköping University Electronic Press

CY - Linköping

T2 - NODALIDA 2015: The 20th Nordic Conference of Computational Linguistics

Y2 - 11 May 2015 through 13 May 2015

ER -

Zapata J, Kirkedal AS. Assessing the Performance of Automatic Speech Recognition Systems When Used by Native and Non-Native Speakers of Three Major Languages in Dictation Workflows. In Megyesi B, editor, Proceedings of the 20th Nordic Conference of Computational Linguistics, NODALIDA 2015. Linköping: Linköping University Electronic Press. 2015. p. 201-210. (NEALT (Northern European Association of Language Technology) Proceedings Series, Vol. 23).