Exploring Graph-structured Passage Representation for Multi-hop Reading Comprehension with Graph Neural Networks

Abstract

Multi-hop reading comprehension focuses on one type of factoid question, where a system needs to properly integrate multiple pieces of evidence to correctly answer a question. Previous work approximates global evidence with local coreference information, encoding coreference chains with DAG-styled GRU layers within a gated-attention reader. However, coreference is limited in providing information for rich inference. We introduce a new method for better connecting global evidence, which forms more complex graphs compared to DAGs. To perform evidence integration on our graphs, we investigate two recent graph neural networks, namely graph convolutional network (GCN) and graph recurrent network (GRN). Experiments on two standard datasets show that richer global information leads to better answers. Our method performs better than all published results on these datasets.

References in corpus (9)

• Teaching Machines to Read and Comprehend
• Bidirectional Attention Flow for Machine Comprehension
• The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations
• Machine Comprehension Using Match-LSTM and Answer Pointer
• SearchQA: A New Q&A Dataset Augmented with Context from a Search Engine
• ReasoNet: Learning to Stop Reading in Machine Comprehension
• Evidence Aggregation for Answer Re-Ranking in Open-Domain Question Answering
• Quasar: Datasets for Question Answering by Search and Reading
• Multi-Perspective Context Matching for Machine Comprehension

Cited by in corpus (23)

• Knowledge Guided Text Retrieval and Reading for Open Domain Question Answering
• Graph Neural Networks for Natural Language Processing: A Survey
• Multi-hop Question Answering via Reasoning Chains
• BertGCN: Transductive Text Classification by Combining GCN and BERT
• Machine Reading Comprehension: The Role of Contextualized Language Models and Beyond
• Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
• Is Graph Structure Necessary for Multi-hop Question Answering?
• GraphFlow: Exploiting Conversation Flow with Graph Neural Networks for Conversational Machine Comprehension
• GNN-LM: Language Modeling based on Global Contexts via GNN
• Stability and Generalization of Graph Convolutional Neural Networks
• Graph-free Multi-hop Reading Comprehension: A Select-to-Guide Strategy
• Multi-Step Inference for Reasoning Over Paragraphs
• Interpretable Multi-Step Reasoning with Knowledge Extraction on Complex Healthcare Question Answering
• A Survey of Document Grounded Dialogue Systems (DGDS)
• A Heterogeneous Graph with Factual, Temporal and Logical Knowledge for Question Answering Over Dynamic Contexts
• GraphSearchNet: Enhancing GNNs via Capturing Global Dependencies for Semantic Code Search
• SPARQA: Skeleton-based Semantic Parsing for Complex Questions over Knowledge Bases
• Memory Augmented Sequential Paragraph Retrieval for Multi-hop Question Answering
• Multi-Instance Learning for End-to-End Knowledge Base Question Answering
• Tackling Graphical NLP problems with Graph Recurrent Networks
• Fine-tuning Multi-hop Question Answering with Hierarchical Graph Network
• Algorithm for Semantic Network Generation from Texts of Low Resource Languages Such as Kiswahili
• ClueReader: Heterogeneous Graph Attention Network for Multi-hop Machine Reading Comprehension