Question answering (QA) and question generation (QG) are closely related tasks that could improve each other; however, the connection of these two tasks is not well explored in the literature. In this paper, we present two training algorithms for learning better QA and QG models through leveraging one another. The first algorithm extends Generative Adversarial Network (GAN), which selectively incorporates artificially generated instances as additional QA training data. The second algorithm is an extension of dual learning, which incorporates the probabilistic correlation of QA and QG as additional regularization in training objectives. To test the scalability of our algorithms, we conduct experiments on both document based and table based question answering tasks. Results show that both algorithms improve a QA model in terms of accuracy and QG model in terms of BLEU score. Moreover, we find that the performance of a QG model could be easily improved by a QA model via policy gradient, however, directly applying GAN that regards all the generated questions as negative instances could not improve the accuracy of the QA model. Our algorithm that selectively assigns labels to generated questions would bring a performance boost.

中文翻译：

---

问答和问题生成的联合学习

问答 (QA) 和问题生成 (QG) 是密切相关的任务，可以相互改进；然而，这两个任务之间的联系在文献中并没有得到很好的探讨。在本文中，我们提出了两种训练算法，通过相互利用来学习更好的 QA 和 QG 模型。第一种算法扩展了生成对抗网络 (GAN)，它有选择地将人工生成的实例合并为额外的 QA 训练数据。第二种算法是对偶学习的扩展，它结合了 QA 和 QG 的概率相关性作为训练目标中的额外正则化。为了测试我们算法的可扩展性，我们对基于文档和基于表格的问答任务进行了实验。结果表明，两种算法都在准确性方面改进了 QA 模型，在 BLEU 分数方面改进了 QG 模型。此外，我们发现 QA 模型可以通过策略梯度轻松提高 QG 模型的性能，但是，直接应用将所有生成的问题视为负面实例的 GAN 并不能提高 QA 模型的准确性。我们有选择地为生成的问题分配标签的算法将带来性能提升。