Robust Question Answering against Distribution Shifts with Test-Time Adaptation: An Empirical Study

Autor:	Ye, Hai, Ding, Yuyang, Li, Juntao, Ng, Hwee Tou
Rok vydání:	2023
Předmět:	Computer Science - Computation and Language
Druh dokumentu:	Working Paper
Popis:	A deployed question answering (QA) model can easily fail when the test data has a distribution shift compared to the training data. Robustness tuning (RT) methods have been widely studied to enhance model robustness against distribution shifts before model deployment. However, can we improve a model after deployment? To answer this question, we evaluate test-time adaptation (TTA) to improve a model after deployment. We first introduce COLDQA, a unified evaluation benchmark for robust QA against text corruption and changes in language and domain. We then evaluate previous TTA methods on COLDQA and compare them to RT methods. We also propose a novel TTA method called online imitation learning (OIL). Through extensive experiments, we find that TTA is comparable to RT methods, and applying TTA after RT can significantly boost the performance on COLDQA. Our proposed OIL improves TTA to be more robust to variation in hyper-parameters and test distributions over time. Comment: Findings of EMNLP 2022
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2302.04618 Zobrazit plný text záznamu View this record from Arxiv