Machine Comprehension by Text-to-Text Neural Question Generation

By Xingdi Yuan, Tong Wang, Caglar Gulcehre, Alessandro Sordoni, Philip Bachman, Saizheng Zhang, Sandeep Subramanian, Adam Trischler. 2017

194 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Natural Language Processing

Citations in the corpus

Addressing Semantic Drift in Question Generation for Semi-Supervised Question Answering (2019, background): “Yuan et al. (2017) proposed to use a language model’s perplexity ( R PPL ) and a QA model’s accuracy ( R QA ) as two rewards but failed to get signiﬁcant improvement.”
Towards a Better Metric for Evaluating Question Generation Systems (2018, methodology): “Some models (Yuan et al., 2017) feed the generated questions to a QA system and use the performance of the QA system as an indicator of the quality of the questions.”
Paragraph-level Neural Question Generation with Maxout Pointer and Gated Self-attention Networks (2018, methodology): “In addition, our model is more concise compared to most of existing ones, e.g. (Yuan et al., 2017; Song et al., 2017).”
Improving Machine Reading Comprehension with General Reading Strategies (2018, methodology): “Previous methods augment the training data for extractive machine reading comprehension and question answering by randomly reordering words or shuffling sentences (Ding and Zhou, 2018; Li and Zhou, 2018) or generating questions through paraphrasing (Yang et al., 2017; Yuan et al., 2017), which requi”
Unsupervised Question Answering by Cloze Translation (2019, background): “Recently there has been interest in question generation using supervised neural models, many trained to generate questions from c, a pairs in SQuAD (Du et al., 2017; Yuan et al., 2017; Zhao et al., 2018; Du and Cardie, 2018; Hosking and Riedel, 2019)”
Evaluation of Text Generation: A Survey (2020, methodology): “MC has been used for tasks like improving question generation (Yuan et al., 2017; Du et al., 2017) and document summarization (Hermann et al., 2015).”
Reinforced Mnemonic Reader for Machine Comprehension (2017, methodology): “Reinforcement learning has been successfully used to solve a wide variety of problems in the ﬁeld of NLP, including abstractive summarization (Paulus, Xiong, and Socher 2017), question generation (Yuan et al. 2017) and dialogue system (Li et al. 2016).”
Learning to Generate Questions by LearningWhat not to Generate (2019, background): “One of the key applications of question generation is to automatically produce question-answer pairs to enhance machine reading comprehension systems [8, 34, 35, 38].”
Neural Models for Key Phrase Extraction and Question Generation (2018, methodology): “Our setup for this module is identical to (Yuan et al., 2017).”
A Joint Model for Question Answering and Question Generation (2017, background): “Yuan et al. (2017) proposed a similar architecture but in addition improved model performance through policy gradient techniques.”
Leveraging Context Information for Natural Question Generation (2018, background): “They either entirely ignore the target answer (Du et al., 2017), or directly hard-code answer positions (Zhou et al., 2017; Yang et al., 2017; Subramanian et al., 2017; Tang et al., 2017; Wang et al., 2017a; Yuan et al., 2017).”
Modeling question asking using neural program generation (2019, background): “Recently, neural network approaches have also been proposed, including retrieval methods which select the best question from past experience [12] and encoder-decoder frameworks which map visual or linguistic inputs to questions [18, 12, 23, 21].”
Improving Question Generation with Sentence-level Semantic Matching and Answer Position Inferring (2019, background): “With the success of reinforcement learning, (Yuan et al. 2017) propose to combine supervised learning and reinforcement learning together for question generation by using policy gradient after training the model in supervised learning way.”
A Recurrent BERT-based Model for Question Generation (2019, background): “Further, (Tang et al., 2017; Wang et al., 2017) proposed joint models to address QG and question answering as a multi-task learning setting.”
Event Extraction as Machine Reading Comprehension (2020, background): “Note we cannot adopt supervised question generation methods (Duan et al., 2017; Yuan et al., 2017; Elsahar et al., 2018), owing to the lack of aligned question-event pairs.”
Transformer-based End-to-End Question Generation (2020, result): “More recently, other techniques have been proposed in order to perform QG. Reinforcement Learning (RL) have produced consistent results for the task by using policy gradients (Yuan et al., 2017).”