Machine Comprehension by Text-to-Text Neural Question Generation
This work addresses the challenge of enhancing question-answering systems through automated question generation, which is incremental as it builds on existing neural methods and datasets.
The authors tackled the problem of generating natural-language questions from documents conditioned on answers, using a recurrent neural model trained with supervised and reinforcement learning on the SQuAD dataset, resulting in improved question-answering system performance.
We propose a recurrent neural model that generates natural-language questions from documents, conditioned on answers. We show how to train the model using a combination of supervised and reinforcement learning. After teacher forcing for standard maximum likelihood training, we fine-tune the model using policy gradient techniques to maximize several rewards that measure question quality. Most notably, one of these rewards is the performance of a question-answering system. We motivate question generation as a means to improve the performance of question answering systems. Our model is trained and evaluated on the recent question-answering dataset SQuAD.