Weakly supervised part-of-speech tagging using eye-tracking data

Maria Jung Barrett; Joachim Bingel; Frank Keller; Anders Søgaard

Weakly supervised part-of-speech tagging using eye-tracking data

Maria Jung Barrett, Joachim Bingel, Frank Keller, Anders Søgaard

Department of Nordic Studies and Linguistics

32 Citations (Scopus)

Abstract

For many of the world's languages, there are no or very few linguistically annotated resources. On the other hand, raw text, and often also dictionaries, can be harvested from the web for many of these languages, and part-of-speech taggers can be trained with these resources. At the same time, previous research shows that eye-tracking data, which can be obtained without explicit annotation, contains clues to partof-speech information. In this work, we bring these two ideas together and show that given raw text, a dictionary, and eyetracking data obtained from naive participants reading text, we can train a weakly supervised PoS tagger using a secondorder HMM with maximum entropy emissions. The best model use type-level aggregates of eye-tracking data and significantly outperforms a baseline that does not have access to eye-tracking data.

Original language	English
Title of host publication	The 54th Annual Meeting of the Association for Computational Linguistics : Proceedings of the Conference
Number of pages	6
Volume	2
Place of Publication	Berlin
Publisher	Association for Computational Linguistics
Publication date	2016
Pages	579-584
ISBN (Print)	9781945626012
Publication status	Published - 2016
Event	ACL 2016 - Duration: 7 Aug 2016 → 12 Aug 2016

Conference

Conference	ACL 2016
Period	07/08/2016 → 12/08/2016

Access to Document

https://www.aclweb.org/anthology/P/P16/P16-2094.pdfLicence: Other

Cite this

Weakly supervised part-of-speech tagging using eye-tracking data. / Barrett, Maria Jung; Bingel, Joachim; Keller, Frank et al.

The 54th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference. Vol. 2 Berlin : Association for Computational Linguistics, 2016. p. 579-584.

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

Barrett, MJ, Bingel, J, Keller, F & Søgaard, A 2016, Weakly supervised part-of-speech tagging using eye-tracking data. in The 54th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference. vol. 2, Association for Computational Linguistics, Berlin, pp. 579-584, ACL 2016, 07/08/2016. <https://www.aclweb.org/anthology/P/P16/P16-2094.pdf>

@inproceedings{7f6782c6c01843d198ffe43a2403ea55,

title = "Weakly supervised part-of-speech tagging using eye-tracking data",

abstract = "For many of the world's languages, there are no or very few linguistically annotated resources. On the other hand, raw text, and often also dictionaries, can be harvested from the web for many of these languages, and part-of-speech taggers can be trained with these resources. At the same time, previous research shows that eye-tracking data, which can be obtained without explicit annotation, contains clues to partof-speech information. In this work, we bring these two ideas together and show that given raw text, a dictionary, and eyetracking data obtained from naive participants reading text, we can train a weakly supervised PoS tagger using a secondorder HMM with maximum entropy emissions. The best model use type-level aggregates of eye-tracking data and significantly outperforms a baseline that does not have access to eye-tracking data.",

author = "Barrett, {Maria Jung} and Joachim Bingel and Frank Keller and Anders S{\o}gaard",

year = "2016",

language = "English",

isbn = "9781945626012",

volume = "2",

pages = "579--584",

booktitle = "The 54th Annual Meeting of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

note = "ACL 2016 ; Conference date: 07-08-2016 Through 12-08-2016",

}

TY - GEN

T1 - Weakly supervised part-of-speech tagging using eye-tracking data

AU - Barrett, Maria Jung

AU - Bingel, Joachim

AU - Keller, Frank

AU - Søgaard, Anders

PY - 2016

Y1 - 2016

N2 - For many of the world's languages, there are no or very few linguistically annotated resources. On the other hand, raw text, and often also dictionaries, can be harvested from the web for many of these languages, and part-of-speech taggers can be trained with these resources. At the same time, previous research shows that eye-tracking data, which can be obtained without explicit annotation, contains clues to partof-speech information. In this work, we bring these two ideas together and show that given raw text, a dictionary, and eyetracking data obtained from naive participants reading text, we can train a weakly supervised PoS tagger using a secondorder HMM with maximum entropy emissions. The best model use type-level aggregates of eye-tracking data and significantly outperforms a baseline that does not have access to eye-tracking data.

AB - For many of the world's languages, there are no or very few linguistically annotated resources. On the other hand, raw text, and often also dictionaries, can be harvested from the web for many of these languages, and part-of-speech taggers can be trained with these resources. At the same time, previous research shows that eye-tracking data, which can be obtained without explicit annotation, contains clues to partof-speech information. In this work, we bring these two ideas together and show that given raw text, a dictionary, and eyetracking data obtained from naive participants reading text, we can train a weakly supervised PoS tagger using a secondorder HMM with maximum entropy emissions. The best model use type-level aggregates of eye-tracking data and significantly outperforms a baseline that does not have access to eye-tracking data.

M3 - Article in proceedings

SN - 9781945626012

VL - 2

SP - 579

EP - 584

BT - The 54th Annual Meeting of the Association for Computational Linguistics

PB - Association for Computational Linguistics

CY - Berlin

T2 - ACL 2016

Y2 - 7 August 2016 through 12 August 2016

ER -

Weakly supervised part-of-speech tagging using eye-tracking data

Abstract

Conference

Access to Document

Fingerprint

Cite this