Cooperative Self-training of Machine Reading Comprehension

TL;DR

This paper introduces RGX, a cooperative self-training framework that automatically generates question-answer pairs to enhance machine reading comprehension models, reducing reliance on annotated data and outperforming existing state-of-the-art methods.

Contribution

The paper presents a novel self-training framework, RGX, combining question generation and answer extraction to improve question answering without needing annotated datasets.

Findings

RGX outperforms SOTA pretrained models on standard benchmarks.

RGX achieves new SOTA performance with limited model size.

The framework enables training on unannotated text corpora.

Abstract

Pretrained language models have significantly improved the performance of downstream language understanding tasks, including extractive question answering, by providing high-quality contextualized word embeddings. However, training question answering models still requires large amounts of annotated data for specific domains. In this work, we propose a cooperative self-training framework, RGX, for automatically generating more non-trivial question-answer pairs to improve model performance. RGX is built upon a masked answer extraction task with an interactive learning environment containing an answer entity Recognizer, a question Generator, and an answer eXtractor. Given a passage with a masked entity, the generator generates a question around the entity, and the extractor is trained to extract the masked entity with the generated question and raw texts. The framework allows the training of question generation and answering models on any text corpora without annotation. Experiment results show that RGX outperforms the state-of-the-art (SOTA) pretrained language models and transfer learning approaches on standard question-answering benchmarks, and yields the new SOTA performance under given model size and transfer learning settings.