Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks

Vardaan Pahuja; Anirban Laha; Shachar Mirkin; Vikas Raykar; Lili Kotlerman; Guy Lev

doi:10.21437/interspeech.2017-1247

Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks

Vardaan Pahuja, Anirban Laha, Shachar Mirkin, Vikas Raykar, Lili Kotlerman, Guy Lev

IBM

Research output: Contribution to journal › Conference article › peer-review

19 Scopus citations

Abstract

The stream of words produced by Automatic Speech Recognition (ASR) systems is typically devoid of punctuations and formatting. Most natural language processing applications expect segmented and well-formatted texts as input, which is not available in ASR output. This paper proposes a novel technique of jointly modeling multiple correlated tasks such as punctuation and capitalization using bidirectional recurrent neural networks, which leads to improved performance for each of these tasks. This method could be extended for joint modeling of any other correlated sequence labeling tasks.

Original language	English
Pages (from-to)	548-552
Number of pages	5
Journal	Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Volume	2017-August
DOIs	https://doi.org/10.21437/interspeech.2017-1247
State	Published - 2017
Externally published	Yes
Event	18th Annual Conference of the International Speech Communication Association, INTERSPEECH 2017 - Stockholm, Sweden Duration: 20 Aug 2017 → 24 Aug 2017

Bibliographical note

Publisher Copyright:
Copyright © 2017 ISCA.

Access to Document

10.21437/interspeech.2017-1247

Cite this

@article{504fdc7d88c24d6fa41f0656b4f3ad7a,

title = "Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks",

abstract = "The stream of words produced by Automatic Speech Recognition (ASR) systems is typically devoid of punctuations and formatting. Most natural language processing applications expect segmented and well-formatted texts as input, which is not available in ASR output. This paper proposes a novel technique of jointly modeling multiple correlated tasks such as punctuation and capitalization using bidirectional recurrent neural networks, which leads to improved performance for each of these tasks. This method could be extended for joint modeling of any other correlated sequence labeling tasks.",

author = "Vardaan Pahuja and Anirban Laha and Shachar Mirkin and Vikas Raykar and Lili Kotlerman and Guy Lev",

note = "Publisher Copyright: Copyright {\textcopyright} 2017 ISCA.; 18th Annual Conference of the International Speech Communication Association, INTERSPEECH 2017 ; Conference date: 20-08-2017 Through 24-08-2017",

year = "2017",

doi = "10.21437/interspeech.2017-1247",

language = "אנגלית",

volume = "2017-August",

pages = "548--552",

journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",

issn = "2308-457X",

}

Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks. / Pahuja, Vardaan; Laha, Anirban; Mirkin, Shachar et al.
In: Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH, Vol. 2017-August, 2017, p. 548-552.

Research output: Contribution to journal › Conference article › peer-review

TY - JOUR

T1 - Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks

AU - Pahuja, Vardaan

AU - Laha, Anirban

AU - Mirkin, Shachar

AU - Raykar, Vikas

AU - Kotlerman, Lili

AU - Lev, Guy

PY - 2017

Y1 - 2017

N2 - The stream of words produced by Automatic Speech Recognition (ASR) systems is typically devoid of punctuations and formatting. Most natural language processing applications expect segmented and well-formatted texts as input, which is not available in ASR output. This paper proposes a novel technique of jointly modeling multiple correlated tasks such as punctuation and capitalization using bidirectional recurrent neural networks, which leads to improved performance for each of these tasks. This method could be extended for joint modeling of any other correlated sequence labeling tasks.

AB - The stream of words produced by Automatic Speech Recognition (ASR) systems is typically devoid of punctuations and formatting. Most natural language processing applications expect segmented and well-formatted texts as input, which is not available in ASR output. This paper proposes a novel technique of jointly modeling multiple correlated tasks such as punctuation and capitalization using bidirectional recurrent neural networks, which leads to improved performance for each of these tasks. This method could be extended for joint modeling of any other correlated sequence labeling tasks.

UR - http://www.scopus.com/inward/record.url?scp=85032866344&partnerID=8YFLogxK

U2 - 10.21437/interspeech.2017-1247

DO - 10.21437/interspeech.2017-1247

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.conferencearticle???

AN - SCOPUS:85032866344

SN - 2308-457X

VL - 2017-August

SP - 548

EP - 552

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

T2 - 18th Annual Conference of the International Speech Communication Association, INTERSPEECH 2017

Y2 - 20 August 2017 through 24 August 2017

ER -

Joint learning of correlated sequence labeling tasks using bidirectional recurrent neural networks

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this