IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection

Rashel Fam, Yves Lepage

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

1 Citation (Scopus)

Abstract

This paper presents the system submitted by IPS-WASEDA University for CoNLL–SIGMORPHON 2018 Shared Task 1: Type level inflection. We develop a system based on a holistic approach which considers whole-word form as a unit, instead of breaking them into smaller pieces (e,g. morphemes) like the baseline systems does. We also implement an encoder-decoder model which has recently become the new standard in many natural language processing (NLP) tasks. The results show that the neural approach outperforms the baseline and our holistic approach on bigger resources settings. The use of data augmentation helps to improve the performance of the model in lower resources settings, although it still cannot beat the other systems. In the end, for the low resources setting, our holistic approach performs best in comparison to the baseline and the neural approach (even with data augmentation).

Original languageEnglish
Title of host publicationCoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task
Subtitle of host publicationUniversal Morphological Reinflection
PublisherAssociation for Computational Linguistics (ACL)
Pages33-42
Number of pages10
ISBN (Electronic)9781948087834
Publication statusPublished - 2018 Jan 1
Event2018 CoNLL-SIGMORPHON Shared Task: Universal Morphological Reinflection, CoNLL 2018 - Brussels, Belgium
Duration: 2018 Oct 31 → …

Publication series

NameCoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection

Conference

Conference2018 CoNLL-SIGMORPHON Shared Task: Universal Morphological Reinflection, CoNLL 2018
CountryBelgium
CityBrussels
Period18/10/31 → …

Fingerprint

holistic approach
resources
Processing
language
performance

ASJC Scopus subject areas

  • Linguistics and Language
  • Artificial Intelligence
  • Human-Computer Interaction

Cite this

Fam, R., & Lepage, Y. (2018). IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection. In CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection (pp. 33-42). (CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection). Association for Computational Linguistics (ACL).

IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection. / Fam, Rashel; Lepage, Yves.

CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection. Association for Computational Linguistics (ACL), 2018. p. 33-42 (CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection).

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

Fam, R & Lepage, Y 2018, IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection. in CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection. CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection, Association for Computational Linguistics (ACL), pp. 33-42, 2018 CoNLL-SIGMORPHON Shared Task: Universal Morphological Reinflection, CoNLL 2018, Brussels, Belgium, 18/10/31.
Fam R, Lepage Y. IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection. In CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection. Association for Computational Linguistics (ACL). 2018. p. 33-42. (CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection).
Fam, Rashel ; Lepage, Yves. / IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection. CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection. Association for Computational Linguistics (ACL), 2018. pp. 33-42 (CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection).
@inproceedings{e1914edf259d4714998c15a285a131bc,
title = "IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection",
abstract = "This paper presents the system submitted by IPS-WASEDA University for CoNLL–SIGMORPHON 2018 Shared Task 1: Type level inflection. We develop a system based on a holistic approach which considers whole-word form as a unit, instead of breaking them into smaller pieces (e,g. morphemes) like the baseline systems does. We also implement an encoder-decoder model which has recently become the new standard in many natural language processing (NLP) tasks. The results show that the neural approach outperforms the baseline and our holistic approach on bigger resources settings. The use of data augmentation helps to improve the performance of the model in lower resources settings, although it still cannot beat the other systems. In the end, for the low resources setting, our holistic approach performs best in comparison to the baseline and the neural approach (even with data augmentation).",
author = "Rashel Fam and Yves Lepage",
year = "2018",
month = "1",
day = "1",
language = "English",
series = "CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection",
publisher = "Association for Computational Linguistics (ACL)",
pages = "33--42",
booktitle = "CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task",

}

TY - GEN

T1 - IPS-WASEDA system at CoNLL–SIGMORPHON 2018 shared task on morphological inflection

AU - Fam, Rashel

AU - Lepage, Yves

PY - 2018/1/1

Y1 - 2018/1/1

N2 - This paper presents the system submitted by IPS-WASEDA University for CoNLL–SIGMORPHON 2018 Shared Task 1: Type level inflection. We develop a system based on a holistic approach which considers whole-word form as a unit, instead of breaking them into smaller pieces (e,g. morphemes) like the baseline systems does. We also implement an encoder-decoder model which has recently become the new standard in many natural language processing (NLP) tasks. The results show that the neural approach outperforms the baseline and our holistic approach on bigger resources settings. The use of data augmentation helps to improve the performance of the model in lower resources settings, although it still cannot beat the other systems. In the end, for the low resources setting, our holistic approach performs best in comparison to the baseline and the neural approach (even with data augmentation).

AB - This paper presents the system submitted by IPS-WASEDA University for CoNLL–SIGMORPHON 2018 Shared Task 1: Type level inflection. We develop a system based on a holistic approach which considers whole-word form as a unit, instead of breaking them into smaller pieces (e,g. morphemes) like the baseline systems does. We also implement an encoder-decoder model which has recently become the new standard in many natural language processing (NLP) tasks. The results show that the neural approach outperforms the baseline and our holistic approach on bigger resources settings. The use of data augmentation helps to improve the performance of the model in lower resources settings, although it still cannot beat the other systems. In the end, for the low resources setting, our holistic approach performs best in comparison to the baseline and the neural approach (even with data augmentation).

UR - http://www.scopus.com/inward/record.url?scp=85072888403&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072888403&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85072888403

T3 - CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection

SP - 33

EP - 42

BT - CoNLL 2018 - Proceedings of the CoNLL-SIGMORPHON 2018 Shared Task

PB - Association for Computational Linguistics (ACL)

ER -