Multilingual Gradient Word-Order Typology from Universal Dependencies

Emi Baylor; Esther Ploeger; Johannes Bjerva

doi:10.48550/arxiv.2402.01513

Multilingual Gradient Word-Order Typology from Universal Dependencies

Emi Baylor^*, Esther Ploeger^*, Johannes Bjerva

^*Corresponding author for this work

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

Abstract

While information from the field of linguistic typology has the potential to improve performance on NLP tasks, reliable typological data is a prerequisite. Existing typological databases, including WALS and Grambank, suffer from inconsistencies primarily caused by their categorical format. Furthermore, typological categorisations by definition differ significantly from the continuous nature of phenomena, as found in natural language corpora. In this paper, we introduce a new seed dataset made up of continuous-valued data, rather than categorical data, that can better reflect the variability of language. While this initial dataset focuses on word-order typology, we also present the methodology used to create the dataset, which can be easily adapted to generate data for a broader set of features and languages.

Original language	English
Title of host publication	Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics : EACL
Volume	2
Place of Publication	St. Julian’s, Malta
Publisher	Association for Computational Linguistics
Publication date	17 Mar 2024
ISBN (Electronic)	979-8-89176-093-6
DOIs	https://doi.org/10.48550/arxiv.2402.01513
Publication status	Published - 17 Mar 2024
Event	The 18th Conference of the European Chapter of the Association for Computational Linguistics - Radisson Blu, St. Julian's, Malta Duration: 17 Mar 2024 → 22 Mar 2024 https://2024.eacl.org/

Conference

Conference	The 18th Conference of the European Chapter of the Association for Computational Linguistics
Location	Radisson Blu
Country/Territory	Malta
City	St. Julian's
Period	17/03/2024 → 22/03/2024
Internet address	https://2024.eacl.org/

Keywords

NLP
Typology

Access to Document

10.48550/arxiv.2402.01513Licence: CC BY 4.0

https://aclanthology.org/2024.eacl-short.6.pdf

AUB Link

Search for the material in Aalborg University Library's search engine

Multilingual Modelling for Resource-Poor Languages
Bjerva, J., Lent, H. C., Chen, Y., Ploeger, E., Fekete, M. R. & Lavrinovics, E.
Carlsbergfondet, Google
01/09/2022 → 31/08/2025
Project: Research

Multilingual Gradient Word-Order Typology from Universal Dependencies
Esther Ploeger (Lecturer)
19 Mar 2024
Activity: Talks and presentations › Conference presentations

Cite this

@inproceedings{73cb14253cc143dbae872f8fafae71ee,

title = "Multilingual Gradient Word-Order Typology from Universal Dependencies",

abstract = "While information from the field of linguistic typology has the potential to improve performance on NLP tasks, reliable typological data is a prerequisite. Existing typological databases, including WALS and Grambank, suffer from inconsistencies primarily caused by their categorical format. Furthermore, typological categorisations by definition differ significantly from the continuous nature of phenomena, as found in natural language corpora. In this paper, we introduce a new seed dataset made up of continuous-valued data, rather than categorical data, that can better reflect the variability of language. While this initial dataset focuses on word-order typology, we also present the methodology used to create the dataset, which can be easily adapted to generate data for a broader set of features and languages.",

keywords = "NLP, Typology",

author = "Emi Baylor and Esther Ploeger and Johannes Bjerva",

year = "2024",

month = mar,

day = "17",

doi = "10.48550/arxiv.2402.01513",

language = "English",

volume = "2",

booktitle = "Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

address = "United States",

note = "The 18th Conference of the European Chapter of the Association for Computational Linguistics, EACL24 ; Conference date: 17-03-2024 Through 22-03-2024",

url = "https://2024.eacl.org/",

}

Baylor, E, Ploeger, E & Bjerva, J 2024, Multilingual Gradient Word-Order Typology from Universal Dependencies. in Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: EACL. vol. 2, Association for Computational Linguistics, St. Julian’s, Malta, The 18th Conference of the European Chapter of the Association for Computational Linguistics, St. Julian's, Malta, 17/03/2024. https://doi.org/10.48550/arxiv.2402.01513

Multilingual Gradient Word-Order Typology from Universal Dependencies. / Baylor, Emi; Ploeger, Esther ; Bjerva, Johannes.
Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: EACL. Vol. 2 St. Julian’s, Malta: Association for Computational Linguistics, 2024.

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

TY - GEN

T1 - Multilingual Gradient Word-Order Typology from Universal Dependencies

AU - Baylor, Emi

AU - Ploeger, Esther

AU - Bjerva, Johannes

PY - 2024/3/17

Y1 - 2024/3/17

N2 - While information from the field of linguistic typology has the potential to improve performance on NLP tasks, reliable typological data is a prerequisite. Existing typological databases, including WALS and Grambank, suffer from inconsistencies primarily caused by their categorical format. Furthermore, typological categorisations by definition differ significantly from the continuous nature of phenomena, as found in natural language corpora. In this paper, we introduce a new seed dataset made up of continuous-valued data, rather than categorical data, that can better reflect the variability of language. While this initial dataset focuses on word-order typology, we also present the methodology used to create the dataset, which can be easily adapted to generate data for a broader set of features and languages.

AB - While information from the field of linguistic typology has the potential to improve performance on NLP tasks, reliable typological data is a prerequisite. Existing typological databases, including WALS and Grambank, suffer from inconsistencies primarily caused by their categorical format. Furthermore, typological categorisations by definition differ significantly from the continuous nature of phenomena, as found in natural language corpora. In this paper, we introduce a new seed dataset made up of continuous-valued data, rather than categorical data, that can better reflect the variability of language. While this initial dataset focuses on word-order typology, we also present the methodology used to create the dataset, which can be easily adapted to generate data for a broader set of features and languages.

KW - NLP

KW - Typology

U2 - 10.48550/arxiv.2402.01513

DO - 10.48550/arxiv.2402.01513

M3 - Article in proceeding

VL - 2

BT - Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics

PB - Association for Computational Linguistics

CY - St. Julian’s, Malta

T2 - The 18th Conference of the European Chapter of the Association for Computational Linguistics

Y2 - 17 March 2024 through 22 March 2024

ER -

Multilingual Gradient Word-Order Typology from Universal Dependencies

Abstract

Conference

Keywords

Access to Document

AUB Link

Fingerprint

Projects

Multilingual Modelling for Resource-Poor Languages

Activities

Multilingual Gradient Word-Order Typology from Universal Dependencies

Cite this