End-to-End QA on COVID-19: Domain Adaptation with Synthetic Training
This work provides a method for improving end-to-end QA performance for researchers and practitioners working with specialized, data-scarce domains like medical literature, which is an incremental improvement.
This paper addresses the challenge of end-to-end question answering (QA) on specialized domains like COVID-19, where neural information retrieval (IR) systems underperform traditional methods. The authors improve performance by using synthetically generated QA examples for domain adaptation, resulting in significant improvements on the CORD-19 collection compared to a state-of-the-art open-domain QA baseline.
End-to-end question answering (QA) requires both information retrieval (IR) over a large document collection and machine reading comprehension (MRC) on the retrieved passages. Recent work has successfully trained neural IR systems using only supervised question answering (QA) examples from open-domain datasets. However, despite impressive performance on Wikipedia, neural IR lags behind traditional term matching approaches such as BM25 in more specific and specialized target domains such as COVID-19. Furthermore, given little or no labeled data, effective adaptation of QA systems can also be challenging in such target domains. In this work, we explore the application of synthetically generated QA examples to improve performance on closed-domain retrieval and MRC. We combine our neural IR and MRC systems and show significant improvements in end-to-end QA on the CORD-19 collection over a state-of-the-art open-domain QA baseline.