DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic

Hariram Veeramani; Surendrabikram Thapa; Usman Naseem

doi:10.18653/v1/2023.arabicnlp-1.63

DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic

Hariram Veeramani, Surendrabikram Thapa, Usman Naseem

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

5 Citations (Scopus)

Abstract

With approximately 400 million speakers worldwide, Arabic ranks as the fifth most-spoken language globally, necessitating advancements in natural language processing. This paper describes the approaches employed for the subtasks outlined in the Nuanced Arabic Dialect Identification (NADI) task at EMNLP 2023. We employ an ensemble of two Arabic language models for the first subtask involving closed country-level dialect identification classification. Similarly, for the second subtask, focused on closed dialect to Modern Standard Arabic (MSA) machine translation, our approach combines sequence-to-sequence models trained on an Arabic-specific dataset. Our team ranks 10th and 3rd on subtask 1 and subtask 2, respectively.

Original language	English
Title of host publication	Proceedings of ArabicNLP 2023
Place of Publication	Stroudsburg
Publisher	Association for Computational Linguistics (ACL)
Pages	614-619
Number of pages	6
ISBN (Electronic)	9781959429272
DOIs	https://doi.org/10.18653/v1/2023.arabicnlp-1.63
Publication status	Published - 2023
Externally published	Yes
Event	1st Arabic Natural Language Processing Conference, ArabicNLP 2023 - Hybrid, Singapore, Singapore Duration: 7 Dec 2023 → 7 Dec 2023

Conference

Conference	1st Arabic Natural Language Processing Conference, ArabicNLP 2023
Country/Territory	Singapore
City	Hybrid, Singapore
Period	7/12/23 → 7/12/23

Access to Document

10.18653/v1/2023.arabicnlp-1.63Licence: Unspecified

Cite this

@inproceedings{ec935a01e14849bbafe41220cf32727e,

title = "DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic",

abstract = "With approximately 400 million speakers worldwide, Arabic ranks as the fifth most-spoken language globally, necessitating advancements in natural language processing. This paper describes the approaches employed for the subtasks outlined in the Nuanced Arabic Dialect Identification (NADI) task at EMNLP 2023. We employ an ensemble of two Arabic language models for the first subtask involving closed country-level dialect identification classification. Similarly, for the second subtask, focused on closed dialect to Modern Standard Arabic (MSA) machine translation, our approach combines sequence-to-sequence models trained on an Arabic-specific dataset. Our team ranks 10th and 3rd on subtask 1 and subtask 2, respectively.",

author = "Hariram Veeramani and Surendrabikram Thapa and Usman Naseem",

year = "2023",

doi = "10.18653/v1/2023.arabicnlp-1.63",

language = "English",

pages = "614--619",

booktitle = "Proceedings of ArabicNLP 2023",

publisher = "Association for Computational Linguistics (ACL)",

address = "United States",

note = "1st Arabic Natural Language Processing Conference, ArabicNLP 2023 ; Conference date: 07-12-2023 Through 07-12-2023",

}

Veeramani, H, Thapa, S & Naseem, U 2023, DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic. in Proceedings of ArabicNLP 2023. Association for Computational Linguistics (ACL), Stroudsburg, pp. 614-619, 1st Arabic Natural Language Processing Conference, ArabicNLP 2023, Hybrid, Singapore, Singapore, 7/12/23. https://doi.org/10.18653/v1/2023.arabicnlp-1.63

DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic. / Veeramani, Hariram; Thapa, Surendrabikram; Naseem, Usman.
Proceedings of ArabicNLP 2023. Stroudsburg: Association for Computational Linguistics (ACL), 2023. p. 614-619.

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

TY - GEN

T1 - DialectNLU at NADI 2023 shared task

T2 - 1st Arabic Natural Language Processing Conference, ArabicNLP 2023

AU - Veeramani, Hariram

AU - Thapa, Surendrabikram

AU - Naseem, Usman

PY - 2023

Y1 - 2023

N2 - With approximately 400 million speakers worldwide, Arabic ranks as the fifth most-spoken language globally, necessitating advancements in natural language processing. This paper describes the approaches employed for the subtasks outlined in the Nuanced Arabic Dialect Identification (NADI) task at EMNLP 2023. We employ an ensemble of two Arabic language models for the first subtask involving closed country-level dialect identification classification. Similarly, for the second subtask, focused on closed dialect to Modern Standard Arabic (MSA) machine translation, our approach combines sequence-to-sequence models trained on an Arabic-specific dataset. Our team ranks 10th and 3rd on subtask 1 and subtask 2, respectively.

AB - With approximately 400 million speakers worldwide, Arabic ranks as the fifth most-spoken language globally, necessitating advancements in natural language processing. This paper describes the approaches employed for the subtasks outlined in the Nuanced Arabic Dialect Identification (NADI) task at EMNLP 2023. We employ an ensemble of two Arabic language models for the first subtask involving closed country-level dialect identification classification. Similarly, for the second subtask, focused on closed dialect to Modern Standard Arabic (MSA) machine translation, our approach combines sequence-to-sequence models trained on an Arabic-specific dataset. Our team ranks 10th and 3rd on subtask 1 and subtask 2, respectively.

UR - http://www.scopus.com/inward/record.url?scp=85176341930&partnerID=8YFLogxK

U2 - 10.18653/v1/2023.arabicnlp-1.63

DO - 10.18653/v1/2023.arabicnlp-1.63

M3 - Conference proceeding contribution

AN - SCOPUS:85176341930

SP - 614

EP - 619

BT - Proceedings of ArabicNLP 2023

PB - Association for Computational Linguistics (ACL)

CY - Stroudsburg

Y2 - 7 December 2023 through 7 December 2023

ER -

DialectNLU at NADI 2023 shared task: transformer based multitask approach jointly integrating dialect and machine translation tasks in Arabic

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this