Question Answering and Question Generation as Dual Tasks

Abstract

We study the problem of joint question answering (QA) and question generation (QG) in this paper. Our intuition is that QA and QG have intrinsic connections and these two tasks could improve each other. On one side, the QA model judges whether the generated question of a QG model is relevant to the answer. On the other side, the QG model provides the probability of generating a question given the answer, which is a useful evidence that in turn facilitates QA. In this paper we regard QA and QG as dual tasks. We propose a training framework that trains the models of QA and QG simultaneously, and explicitly leverages their probabilistic correlation to guide the training process of both models. We implement a QG model based on sequence-to-sequence learning, and a QA model based on recurrent neural network. As all the components of the QA and QG models are differentiable, all the parameters involved in these two models could be conventionally learned with back propagation. We conduct experiments on three datasets. Empirical results show that our training framework improves both QA and QG tasks. The improved QA model performs comparably with strong baseline approaches on all three datasets.

References in corpus (8)

• ADADELTA: An Adaptive Learning Rate Method
• Convolutional Neural Network Architectures for Matching Natural Language Sentences
• Deep Learning for Answer Sentence Selection
• Attentive Pooling Networks
• Encoding Source Language with Convolutional Neural Network for Machine Translation
• Neural Question Generation from Text: A Preliminary Study
• Semi-Supervised QA with Generative Domain-Adaptive Nets
• Question Generation from a Knowledge Base with Web Exploration

Cited by in corpus (50)

• StoryBuddy: A Human-AI Collaborative Chatbot for Parent-Child Interactive Storytelling with Flexible Parental Involvement
• A Comparative Study of AI-Generated (GPT-4) and Human-crafted MCQs in Programming Education
• Learning to Generate Questions by Learning What not to Generate
• CoaCor: Code Annotation for Code Retrieval with Reinforcement Learning
• Toward Subgraph-Guided Knowledge Graph Question Generation with Graph Neural Networks
• Learning to Answer by Learning to Ask: Getting the Best of GPT-2 and BERT Worlds
• A Unified Query-based Generative Model for Question Generation and Question Answering
• Reinforcement Learning Based Graph-to-Sequence Model for Natural Question Generation
• Multi-hop Question Generation with Graph Convolutional Network
• Automatic question generation based on sentence structure analysis using machine learning approach
• Visual Question Generation as Dual Task of Visual Question Answering
• Embedding-based Zero-shot Retrieval through Query Generation
• End-to-End Synthetic Data Generation for Domain Adaptation of Question Answering Systems
• Energy-based View of Retrosynthesis
• CDL: Curriculum Dual Learning for Emotion-Controllable Response Generation
• OneStop QAMaker: Extract Question-Answer Pairs from Text in a One-Stop Approach
• Addressing Semantic Drift in Question Generation for Semi-Supervised Question Answering
• DAL: Dual Adversarial Learning for Dialogue Generation
• Logician: A Unified End-to-End Neural Approach for Open-Domain Information Extraction
• Improving Neural Question Generation using Answer Separation
• Learning to Ask Questions in Open-domain Conversational Systems with Typed Decoders
• EQG-RACE: Examination-Type Question Generation
• Multi-Task Learning with Language Modeling for Question Generation
• Constructing Emotion Consensus and Utilizing Unpaired Data for Empathetic Dialogue Generation
• Two Birds, One Stone: A Simple, Unified Model for Text Generation from Structured and Unstructured Data
• Jointly Learning Semantic Parser and Natural Language Generator via Dual Information Maximization
• Towards a Better Metric for Evaluating Question Generation Systems
• Learning to Ask Screening Questions for Job Postings
• Decoupled Dialogue Modeling and Semantic Parsing for Multi-Turn Text-to-SQL
• Easy-to-Hard: Leveraging Simple Questions for Complex Question Generation
• A Generative Approach to Question Answering
• Attentive Convolution: Equipping CNNs with RNN-style Attention Mechanisms
• Enhancing Question Generation with Commonsense Knowledge
• Question Generation from SQL Queries Improves Neural Semantic Parsing
• Asking Questions Like Educational Experts: Automatically Generating Question-Answer Pairs on Real-World Examination Data
• Dual Learning: Theoretical Study and an Algorithmic Extension
• Neural-Symbolic Solver for Math Word Problems with Auxiliary Tasks
• Semantic Parsing with Dual Learning
• Review-based Question Generation with Adaptive Instance Transfer and Augmentation
• Stronger Transformers for Neural Multi-Hop Question Generation
• Contrastive Domain Adaptation for Question Answering using Limited Text Corpora
• A Mutual Information Maximization Approach for the Spurious Solution Problem in Weakly Supervised Question Answering
• Back-Training excels Self-Training at Unsupervised Domain Adaptation of Question Generation and Passage Retrieval
• A Question Type Driven and Copy Loss Enhanced Frameworkfor Answer-Agnostic Neural Question Generation
• Generating Highly Relevant Questions
• Learning Dual Retrieval Module for Semi-supervised Relation Extraction
• Semi-Supervised Few-Shot Learning for Dual Question-Answer Extraction
• Question-type Driven Question Generation
• Keeping Notes: Conditional Natural Language Generation with a Scratchpad Mechanism
• More Robust Dense Retrieval with Contrastive Dual Learning