Zero-shot Reading Comprehension by Cross-lingual Transfer Learning with Multi-lingual Language Representation Model

TL;DR

This paper demonstrates that zero-shot cross-lingual transfer learning for reading comprehension is feasible using a multi-lingual language model, without translating data, and explores what the model learns in this setting.

Contribution

It systematically evaluates zero-shot transfer learning for reading comprehension with a multi-lingual model, showing its effectiveness without data translation.

Findings

Zero-shot learning is feasible with pre-trained multi-lingual models.

Translating source data into target language degrades performance.

The model learns transferable representations across languages.

Abstract

Because it is not feasible to collect training data for every language, there is a growing interest in cross-lingual transfer learning. In this paper, we systematically explore zero-shot cross-lingual transfer learning on reading comprehension tasks with a language representation model pre-trained on multi-lingual corpus. The experimental results show that with pre-trained language representation zero-shot learning is feasible, and translating the source data into the target language is not necessary and even degrades the performance. We further explore what does the model learn in zero-shot setting.