On Efficiently Acquiring Annotations for Multilingual Models

TL;DR

This paper demonstrates that joint multilingual learning combined with active learning significantly improves data efficiency and performance across various multilingual NLP tasks compared to traditional methods.

Contribution

The authors introduce a simple, effective approach that combines joint multilingual training with active learning to enhance data efficiency and model performance in multilingual NLP tasks.

Findings

Joint learning outperforms separate language models.

Active learning provides complementary benefits.

Method outperforms traditional approaches under budget constraints.

Abstract

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.