From Masked Language Modeling to Translation: Non-English Auxiliary Tasks Improve Zero-shot Spoken Language Understanding

TL;DR

This paper introduces xSID, a new benchmark for cross-lingual spoken language understanding in 13 languages, and proposes joint learning methods using auxiliary tasks like masked language modeling and translation to improve low-resource language performance.

Contribution

It presents a new multilingual SLU benchmark and demonstrates effective joint learning strategies with auxiliary tasks for low-resource language transfer.

Findings

Joint learning with masked language modeling improves slot detection.

Machine translation transfer enhances intent classification.

The approach benefits low-resource language SLU performance.

Abstract

The lack of publicly available evaluation data for low-resource languages limits progress in Spoken Language Understanding (SLU). As key tasks like intent classification and slot filling require abundant training data, it is desirable to reuse existing data in high-resource languages to develop models for low-resource scenarios. We introduce xSID, a new benchmark for cross-lingual Slot and Intent Detection in 13 languages from 6 language families, including a very low-resource dialect. To tackle the challenge, we propose a joint learning approach, with English SLU training data and non-English auxiliary tasks from raw text, syntax and translation for transfer. We study two setups which differ by type and language coverage of the pre-trained embeddings. Our results show that jointly learning the main tasks with masked language modeling is effective for slots, while machine translation transfer works best for intent classification.