Multilingual Transfer Learning for QA using Translation as Data Augmentation

Authors

  • Mihaela Bornea IBM Research
  • Lin Pan IBM Research
  • Sara Rosenthal IBM Research
  • Radu Florian IBM Research
  • Avirup Sil IBM Research

Keywords:

Question Answering

Abstract

Prior work on multilingual question answering has mostly focused on using large multilingual pre-trained language models (LM) to perform zero-shot language-wise learning: train a QA model on English and test on other languages. In this work, we explore strategies that improve cross-lingual transfer by bringing the multilingual embeddings closer in the semantic space. Our first strategy augments the original English training data with machine translation-generated data. This results in a corpus of multilingual silver-labeled QA pairs that is 14 times larger than the original training set. In addition, we propose two novel strategies, language adversarial training and language arbitration framework, which significantly improve the (zero-resource) cross-lingual transfer performance and result in LM embeddings that are less language-variant. Empirically, we show that the proposed models outperform the previous zero-shot baseline on the recently introduced multilingual MLQA and TyDiQA datasets.

Downloads

Published

2021-05-18

How to Cite

Bornea, M., Pan, L., Rosenthal, S., Florian, R., & Sil, A. (2021). Multilingual Transfer Learning for QA using Translation as Data Augmentation. Proceedings of the AAAI Conference on Artificial Intelligence, 35(14), 12583-12591. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17491

Issue

Section

AAAI Technical Track on Speech and Natural Language Processing I