On Efficiently Acquiring Annotations for Multilingual Models

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3AWT4LPUIU" target="_blank" >RIV/00216208:11320/22:WT4LPUIU - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2022.acl-short.9" target="_blank" >https://aclanthology.org/2022.acl-short.9</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2022.acl-short.9" target="_blank" >10.18653/v1/2022.acl-short.9</a>

Jazyk výsledku

angličtina

Název v původním jazyce

On Efficiently Acquiring Annotations for Multilingual Models

Popis výsledku v původním jazyce

When tasked with supporting multiple languages for a given problem, two approaches have arisen: training a model for each language with the annotation budget divided equally among them, and training on a high-resource language followed by zero-shot transfer to the remaining languages. In this work, we show that the strategy of joint learning across multiple languages using a single model performs substantially better than the aforementioned alternatives. We also demonstrate that active learning provides additional, complementary benefits. We show that this simple approach enables the model to be data efficient by allowing it to arbitrate its annotation budget to query languages it is less certain on. We illustrate the effectiveness of our proposed method on a diverse set of tasks: a classification task with 4 languages, a sequence tagging task with 4 languages and a dependency parsing task with 5 languages. Our proposed method, whilst simple, substantially outperforms the other viable alternatives for building a model in a multilingual setting under constrained budgets.

Název v anglickém jazyce

On Efficiently Acquiring Annotations for Multilingual Models

Popis výsledku anglicky

When tasked with supporting multiple languages for a given problem, two approaches have arisen: training a model for each language with the annotation budget divided equally among them, and training on a high-resource language followed by zero-shot transfer to the remaining languages. In this work, we show that the strategy of joint learning across multiple languages using a single model performs substantially better than the aforementioned alternatives. We also demonstrate that active learning provides additional, complementary benefits. We show that this simple approach enables the model to be data efficient by allowing it to arbitrate its annotation budget to query languages it is less certain on. We illustrate the effectiveness of our proposed method on a diverse set of tasks: a classification task with 4 languages, a sequence tagging task with 4 languages and a dependency parsing task with 5 languages. Our proposed method, whilst simple, substantially outperforms the other viable alternatives for building a model in a multilingual setting under constrained budgets.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

ISBN

978-1-955917-22-3

ISSN

—

e-ISSN

—

Počet stran výsledku

17

Strana od-do

69-85

Název nakladatele

Association for Computational Linguistics

Místo vydání

—

Místo konání akce

Dublin, Ireland

Datum konání akce

1. 1. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—