XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering

TL;DR

XLDA is a cross-lingual data augmentation technique that improves multilingual NLP tasks by replacing input segments with translations, leading to state-of-the-art results on several benchmarks.

Contribution

The paper introduces XLDA, a novel data augmentation method that significantly enhances multilingual NLP performance, especially for low-resource languages.

Findings

XLDA improves performance on 14 languages in XNLI benchmark.

Achieves up to 4.8% improvement, setting new state-of-the-art for Greek, Turkish, and Urdu.

Provides a 1.0% increase on SQuAD question answering task.

Abstract

While natural language processing systems often focus on a single language, multilingual transfer learning has the potential to improve performance, especially for low-resource languages. We introduce XLDA, cross-lingual data augmentation, a method that replaces a segment of the input text with its translation in another language. XLDA enhances performance of all 14 tested languages of the cross-lingual natural language inference (XNLI) benchmark. With improvements of up to $4.8\%$, training with XLDA achieves state-of-the-art performance for Greek, Turkish, and Urdu. XLDA is in contrast to, and performs markedly better than, a more naive approach that aggregates examples in various languages in a way that each example is solely in one language. On the SQuAD question answering task, we see that XLDA provides a $1.0\%$ performance increase on the English evaluation set. Comprehensive experiments suggest that most languages are effective as cross-lingual augmentors, that XLDA is robust to a wide range of translation quality, and that XLDA is even more effective for randomly initialized models than for pretrained models.