Cascading Adaptors to Leverage English Data to Improve Performance of Question Answering for Low-Resource Languages

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Pandya, Hariom A., Ardeshna, Bhavik, Bhatt, Brijesh S.
Natura: Preprint
Pubblicazione: 2021
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914932527726592
author Pandya, Hariom A.
Ardeshna, Bhavik
Bhatt, Brijesh S.
author_facet Pandya, Hariom A.
Ardeshna, Bhavik
Bhatt, Brijesh S.
contents Transformer based architectures have shown notable results on many down streaming tasks including question answering. The availability of data, on the other hand, impedes obtaining legitimate performance for low-resource languages. In this paper, we investigate the applicability of pre-trained multilingual models to improve the performance of question answering in low-resource languages. We tested four combinations of language and task adapters using multilingual transformer architectures on seven languages similar to MLQA dataset. Additionally, we have also proposed zero-shot transfer learning of low-resource question answering using language and task adapters. We observed that stacking the language and the task adapters improves the multilingual transformer models' performance significantly for low-resource languages.
format Preprint
id arxiv_https___arxiv_org_abs_2112_09866
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Cascading Adaptors to Leverage English Data to Improve Performance of Question Answering for Low-Resource Languages
Pandya, Hariom A.
Ardeshna, Bhavik
Bhatt, Brijesh S.
Computation and Language
Artificial Intelligence
Human-Computer Interaction
Information Retrieval
Machine Learning
Transformer based architectures have shown notable results on many down streaming tasks including question answering. The availability of data, on the other hand, impedes obtaining legitimate performance for low-resource languages. In this paper, we investigate the applicability of pre-trained multilingual models to improve the performance of question answering in low-resource languages. We tested four combinations of language and task adapters using multilingual transformer architectures on seven languages similar to MLQA dataset. Additionally, we have also proposed zero-shot transfer learning of low-resource question answering using language and task adapters. We observed that stacking the language and the task adapters improves the multilingual transformer models' performance significantly for low-resource languages.
title Cascading Adaptors to Leverage English Data to Improve Performance of Question Answering for Low-Resource Languages
topic Computation and Language
Artificial Intelligence
Human-Computer Interaction
Information Retrieval
Machine Learning
url https://arxiv.org/abs/2112.09866