Syllable-aware Neural Language Models: A Failure to Beat Character-aware Ones

Zhenisbek Assylbekov, Rustem Takhanov, Bagdat Myrzakhmetov, Jonathan N. Washington

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

3 Citations (Scopus)

Abstract

Syllabification does not seem to improve word-level RNN language modeling quality when compared to character-based segmentation. However, our best syllable-aware language model, achieving performance comparable to the competitive character-aware model, has 18%–33% fewer parameters and is trained 1.2–2.2 times faster.

Original languageEnglish
Title of host publicationEMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings
PublisherAssociation for Computational Linguistics (ACL)
Pages1866-1872
Number of pages7
ISBN (Electronic)9781945626838
Publication statusPublished - Jan 1 2017
Event2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017 - Copenhagen, Denmark
Duration: Sep 9 2017Sep 11 2017

Publication series

NameEMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

Conference

Conference2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017
CountryDenmark
CityCopenhagen
Period9/9/179/11/17

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems
  • Computational Theory and Mathematics

Cite this

Assylbekov, Z., Takhanov, R., Myrzakhmetov, B., & Washington, J. N. (2017). Syllable-aware Neural Language Models: A Failure to Beat Character-aware Ones. In EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings (pp. 1866-1872). (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings). Association for Computational Linguistics (ACL).

Syllable-aware Neural Language Models : A Failure to Beat Character-aware Ones. / Assylbekov, Zhenisbek; Takhanov, Rustem; Myrzakhmetov, Bagdat; Washington, Jonathan N.

EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2017. p. 1866-1872 (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).

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

Assylbekov, Z, Takhanov, R, Myrzakhmetov, B & Washington, JN 2017, Syllable-aware Neural Language Models: A Failure to Beat Character-aware Ones. in EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings, Association for Computational Linguistics (ACL), pp. 1866-1872, 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, Copenhagen, Denmark, 9/9/17.
Assylbekov Z, Takhanov R, Myrzakhmetov B, Washington JN. Syllable-aware Neural Language Models: A Failure to Beat Character-aware Ones. In EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL). 2017. p. 1866-1872. (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
Assylbekov, Zhenisbek ; Takhanov, Rustem ; Myrzakhmetov, Bagdat ; Washington, Jonathan N. / Syllable-aware Neural Language Models : A Failure to Beat Character-aware Ones. EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2017. pp. 1866-1872 (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
@inproceedings{1ed6c4a5b008465fb301df1d285b1c59,
title = "Syllable-aware Neural Language Models: A Failure to Beat Character-aware Ones",
abstract = "Syllabification does not seem to improve word-level RNN language modeling quality when compared to character-based segmentation. However, our best syllable-aware language model, achieving performance comparable to the competitive character-aware model, has 18{\%}–33{\%} fewer parameters and is trained 1.2–2.2 times faster.",
author = "Zhenisbek Assylbekov and Rustem Takhanov and Bagdat Myrzakhmetov and Washington, {Jonathan N.}",
year = "2017",
month = "1",
day = "1",
language = "English",
series = "EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings",
publisher = "Association for Computational Linguistics (ACL)",
pages = "1866--1872",
booktitle = "EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings",

}

TY - GEN

T1 - Syllable-aware Neural Language Models

T2 - A Failure to Beat Character-aware Ones

AU - Assylbekov, Zhenisbek

AU - Takhanov, Rustem

AU - Myrzakhmetov, Bagdat

AU - Washington, Jonathan N.

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Syllabification does not seem to improve word-level RNN language modeling quality when compared to character-based segmentation. However, our best syllable-aware language model, achieving performance comparable to the competitive character-aware model, has 18%–33% fewer parameters and is trained 1.2–2.2 times faster.

AB - Syllabification does not seem to improve word-level RNN language modeling quality when compared to character-based segmentation. However, our best syllable-aware language model, achieving performance comparable to the competitive character-aware model, has 18%–33% fewer parameters and is trained 1.2–2.2 times faster.

UR - http://www.scopus.com/inward/record.url?scp=85060581134&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85060581134&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85060581134

T3 - EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

SP - 1866

EP - 1872

BT - EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

PB - Association for Computational Linguistics (ACL)

ER -