The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline

Yves Scherrer; Aleksandra Miletic Haddad; Olli Kuparinen

doi:10.18653/v1/2023.arabicnlp-1.73

The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline

Yves Scherrer, Aleksandra Miletic Haddad, Olli Kuparinen

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

Abstract

The Helsinki-NLP team participated in the NADI 2023 shared tasks on Arabic dialect translation with seven submissions. We used statistical (SMT) and neural machine translation (NMT) methods and explored character- and subword-based data preprocessing. Our submissions placed second in both tracks. In the open track, our winning submission is a character-level SMT system with additional Modern Standard Arabic language models. In the closed track, our best BLEU scores were obtained with the leave-as-is baseline, a simple copy of the input, and narrowly followed by SMT systems. In both tracks, fine-tuning existing multilingual models such as AraT5 or ByT5 did not yield superior performance compared to SMT.

Original language	English
Title of host publication	Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023)
Editors	Hassan Sawaf, Samhaa El-Beltagy, Wajdi Zaghouani, Walid Magdy, Ahmed Abdelali, Nadi Tomeh, Ibrahim Abu Farha, Nizar Habash, Salam Khalifa, Amr Keleg, Hatem Haddad, Imed Zitouni, Khalil Mrini, Rawan Almatham
Number of pages	8
Place of Publication	Stroudsburg
Publisher	The Association for Computational Linguistics
Publication date	1 Dec 2023
Pages	670-677
ISBN (Electronic)	978-1-959429-27-2
DOIs	https://doi.org/10.18653/v1/2023.arabicnlp-1.73
Publication status	Published - 1 Dec 2023
MoE publication type	A4 Article in conference proceedings
Event	Arabic Natural Language Processing Conference - , Singapore Duration: 7 Dec 2023 → 7 Dec 2023 https://arabicnlp2023.sigarab.org/

Fields of Science

113 Computer and information sciences
6121 Languages

Access to Document

10.18653/v1/2023.arabicnlp-1.73Licence: CC BY

2023.arabicnlp-1.73Final published version, 145 KBLicence: CC BY

CorCoDial: Corpus-based computational dialectology: exploiting machine translation techniques to extract, visualize and interpret dialectal patterns
Scherrer, Y. (Project manager), Tiedemann, J. (Project manager), Kuparinen, O. V. (Participant), Mickus, T. (Participant), Miletic Haddad, A. (Participant), Psaltaki, E. (Participant), Roemling, D. (Participant), Siewert, J. (Participant) & Siewert, J. (Participant)
Suomen Akatemia Projektilaskutus
01/09/2021 → 31/08/2025
Project: Research Council of Finland: Academy Project

Cite this

Scherrer, Y., Miletic Haddad, A., & Kuparinen, O. (2023). The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline. In H. Sawaf, S. El-Beltagy, W. Zaghouani, W. Magdy, A. Abdelali, N. Tomeh, I. Abu Farha, N. Habash, S. Khalifa, A. Keleg, H. Haddad, I. Zitouni, K. Mrini, & R. Almatham (Eds.), Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023) (pp. 670-677). The Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.arabicnlp-1.73

Scherrer, Yves ; Miletic Haddad, Aleksandra ; Kuparinen, Olli. / The Helsinki-NLP Submissions at NADI 2023 Shared Task : Walking the Baseline. Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023). editor / Hassan Sawaf ; Samhaa El-Beltagy ; Wajdi Zaghouani ; Walid Magdy ; Ahmed Abdelali ; Nadi Tomeh ; Ibrahim Abu Farha ; Nizar Habash ; Salam Khalifa ; Amr Keleg ; Hatem Haddad ; Imed Zitouni ; Khalil Mrini ; Rawan Almatham. Stroudsburg : The Association for Computational Linguistics, 2023. pp. 670-677

@inproceedings{cd83c895557c4931b016d871e339febf,

title = "The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline",

abstract = "The Helsinki-NLP team participated in the NADI 2023 shared tasks on Arabic dialect translation with seven submissions. We used statistical (SMT) and neural machine translation (NMT) methods and explored character- and subword-based data preprocessing. Our submissions placed second in both tracks. In the open track, our winning submission is a character-level SMT system with additional Modern Standard Arabic language models. In the closed track, our best BLEU scores were obtained with the leave-as-is baseline, a simple copy of the input, and narrowly followed by SMT systems. In both tracks, fine-tuning existing multilingual models such as AraT5 or ByT5 did not yield superior performance compared to SMT.",

keywords = "113 Computer and information sciences, 6121 Languages",

author = "Yves Scherrer and {Miletic Haddad}, Aleksandra and Olli Kuparinen",

year = "2023",

month = dec,

day = "1",

doi = "10.18653/v1/2023.arabicnlp-1.73",

language = "English",

pages = "670--677",

editor = "Hassan Sawaf and Samhaa El-Beltagy and Wajdi Zaghouani and Walid Magdy and Ahmed Abdelali and Nadi Tomeh and {Abu Farha}, Ibrahim and Nizar Habash and Salam Khalifa and Amr Keleg and Hatem Haddad and Imed Zitouni and Khalil Mrini and Rawan Almatham",

booktitle = "Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023)",

publisher = "The Association for Computational Linguistics",

address = "United States",

note = "Arabic Natural Language Processing Conference, ArabicNLP 2023 ; Conference date: 07-12-2023 Through 07-12-2023",

url = "https://arabicnlp2023.sigarab.org/",

}

Scherrer, Y , Miletic Haddad, A & Kuparinen, O 2023, The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline. in H Sawaf, S El-Beltagy, W Zaghouani, W Magdy, A Abdelali, N Tomeh, I Abu Farha, N Habash, S Khalifa, A Keleg, H Haddad, I Zitouni, K Mrini & R Almatham (eds), Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023). The Association for Computational Linguistics, Stroudsburg, pp. 670-677, Arabic Natural Language Processing Conference, Singapore, 07/12/2023. https://doi.org/10.18653/v1/2023.arabicnlp-1.73

The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline. / Scherrer, Yves ; Miletic Haddad, Aleksandra; Kuparinen, Olli.
Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023). ed. / Hassan Sawaf; Samhaa El-Beltagy; Wajdi Zaghouani; Walid Magdy; Ahmed Abdelali; Nadi Tomeh; Ibrahim Abu Farha; Nizar Habash; Salam Khalifa; Amr Keleg; Hatem Haddad; Imed Zitouni; Khalil Mrini; Rawan Almatham. Stroudsburg: The Association for Computational Linguistics, 2023. p. 670-677.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - The Helsinki-NLP Submissions at NADI 2023 Shared Task

T2 - Arabic Natural Language Processing Conference

AU - Scherrer, Yves

AU - Miletic Haddad, Aleksandra

AU - Kuparinen, Olli

PY - 2023/12/1

Y1 - 2023/12/1

N2 - The Helsinki-NLP team participated in the NADI 2023 shared tasks on Arabic dialect translation with seven submissions. We used statistical (SMT) and neural machine translation (NMT) methods and explored character- and subword-based data preprocessing. Our submissions placed second in both tracks. In the open track, our winning submission is a character-level SMT system with additional Modern Standard Arabic language models. In the closed track, our best BLEU scores were obtained with the leave-as-is baseline, a simple copy of the input, and narrowly followed by SMT systems. In both tracks, fine-tuning existing multilingual models such as AraT5 or ByT5 did not yield superior performance compared to SMT.

AB - The Helsinki-NLP team participated in the NADI 2023 shared tasks on Arabic dialect translation with seven submissions. We used statistical (SMT) and neural machine translation (NMT) methods and explored character- and subword-based data preprocessing. Our submissions placed second in both tracks. In the open track, our winning submission is a character-level SMT system with additional Modern Standard Arabic language models. In the closed track, our best BLEU scores were obtained with the leave-as-is baseline, a simple copy of the input, and narrowly followed by SMT systems. In both tracks, fine-tuning existing multilingual models such as AraT5 or ByT5 did not yield superior performance compared to SMT.

KW - 113 Computer and information sciences

KW - 6121 Languages

U2 - 10.18653/v1/2023.arabicnlp-1.73

DO - 10.18653/v1/2023.arabicnlp-1.73

M3 - Conference contribution

SP - 670

EP - 677

BT - Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023)

A2 - Sawaf, Hassan

A2 - El-Beltagy, Samhaa

A2 - Zaghouani, Wajdi

A2 - Magdy, Walid

A2 - Abdelali, Ahmed

A2 - Tomeh, Nadi

A2 - Abu Farha, Ibrahim

A2 - Habash, Nizar

A2 - Khalifa, Salam

A2 - Keleg, Amr

A2 - Haddad, Hatem

A2 - Zitouni, Imed

A2 - Mrini, Khalil

A2 - Almatham, Rawan

PB - The Association for Computational Linguistics

CY - Stroudsburg

Y2 - 7 December 2023 through 7 December 2023

ER -

Scherrer Y , Miletic Haddad A, Kuparinen O. The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline. In Sawaf H, El-Beltagy S, Zaghouani W, Magdy W, Abdelali A, Tomeh N, Abu Farha I, Habash N, Khalifa S, Keleg A, Haddad H, Zitouni I, Mrini K, Almatham R, editors, Proceedings of the The First Arabic Natural Language Processing Conference (ArabicNLP 2023). Stroudsburg: The Association for Computational Linguistics. 2023. p. 670-677 doi: 10.18653/v1/2023.arabicnlp-1.73

The Helsinki-NLP Submissions at NADI 2023 Shared Task: Walking the Baseline

Abstract

Fields of Science

Access to Document

Projects

CorCoDial: Corpus-based computational dialectology: exploiting machine translation techniques to extract, visualize and interpret dialectal patterns

Cite this