Cooperative Self-training of Machine Reading Comprehension
This addresses the data annotation bottleneck for domain-specific question answering, offering a method to train models on any text corpora without manual labeling.
The authors tackled the problem of training question answering models without requiring large annotated datasets by proposing RGX, a cooperative self-training framework that automatically generates question-answer pairs from raw text. The result showed that RGX outperformed state-of-the-art pretrained language models and transfer learning approaches on standard benchmarks, achieving new SOTA performance under given model size and transfer learning settings.
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.