Back-Training excels Self-Training at Unsupervised Domain Adaptation of Question Generation and Passage Retrieval

Autor:	Kulshreshtha, Devang, Belfer, Robert, Serban, Iulian Vlad, Reddy, Siva
Rok vydání:	2021
Předmět:	Computer Science - Computation and Language Computer Science - Artificial Intelligence Computer Science - Machine Learning
Druh dokumentu:	Working Paper
Popis:	In this work, we introduce back-training, an alternative to self-training for unsupervised domain adaptation (UDA) from source to target domain. While self-training generates synthetic training data where natural inputs are aligned with noisy outputs, back-training results in natural outputs aligned with noisy inputs. This significantly reduces the gap between the target domain and synthetic data distribution, and reduces model overfitting to the source domain. We run UDA experiments on question generation and passage retrieval from the \textit{Natural Questions} domain to machine learning and biomedical domains. We find that back-training vastly outperforms self-training by a mean improvement of 7.8 BLEU-4 points on generation, and 17.6\% top-20 retrieval accuracy across both domains. We further propose consistency filters to remove low-quality synthetic data before training. We also release a new domain-adaptation dataset- \textit{MLQuestions} containing 35K unaligned questions, 50K unaligned passages, and 3K aligned question-passage pairs. Comment: EMNLP 2021
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2104.08801 Zobrazit plný text záznamu View this record from Arxiv