Learning Recurrent Span Representations for Extractive Question Answering

Abstract

The reading comprehension task, that asks questions about a given evidence document, is a central problem in natural language understanding. Recent formulations of this task have typically focused on answer selection from a set of candidates pre-defined manually or through the use of an external NLP pipeline. However, Rajpurkar et al. (2016) recently released the SQuAD dataset in which the answers can be arbitrary strings from the supplied text. In this paper, we focus on this answer extraction task, presenting a novel model architecture that efficiently builds fixed length representations of all spans in the evidence document with a recurrent network. We show that scoring explicit span representations significantly improves performance over other approaches that factor the prediction into separate predictions about words or start and end markers. Our approach improves upon the best published results of Wang & Jiang (2016) by 5% and decreases the error of Rajpurkar et al.'s baseline by > 50%.

References in corpus (4)

• LSTM: A Search Space Odyssey
• Machine Comprehension Using Match-LSTM and Answer Pointer
• A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task
• Dataset and Neural Recurrent Sequence Labeling Model for Open-Domain Factoid Question Answering

Cited by in corpus (41)

• REALM: Retrieval-Augmented Language Model Pre-Training
• QANet: Combining Local Convolution with Global Self-Attention for Reading Comprehension
• ReasoNet: Learning to Stop Reading in Machine Comprehension
• Adversarial Examples for Evaluating Reading Comprehension Systems
• Multi-Perspective Context Matching for Machine Comprehension
• Reading Wikipedia to Answer Open-Domain Questions
• A Unified Query-based Generative Model for Question Generation and Question Answering
• S-Net: From Answer Extraction to Answer Generation for Machine Reading Comprehension
• U-Net: Machine Reading Comprehension with Unanswerable Questions
• Stochastic Answer Networks for Machine Reading Comprehension
• A Joint Training Dual-MRC Framework for Aspect Based Sentiment Analysis
• Efficient and Robust Question Answering from Minimal Context over Documents
• GraphFlow: Exploiting Conversation Flow with Graph Neural Networks for Conversational Machine Comprehension
• Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification
• The Effect of Natural Distribution Shift on Question Answering Models
• Structural Embedding of Syntactic Trees for Machine Comprehension
• Improving Background Based Conversation with Context-aware Knowledge Pre-selection
• Ruminating Reader: Reasoning with Gated Multi-Hop Attention
• Parameter-Efficient Abstractive Question Answering over Tables or Text
• Joint Training of Candidate Extraction and Answer Selection for Reading Comprehension
• End-to-End Neural Event Coreference Resolution
• Two-Stage Synthesis Networks for Transfer Learning in Machine Comprehension
• Making Neural Machine Reading Comprehension Faster
• Learning Dense Representations of Phrases at Scale
• Mapping Natural Language Instructions to Mobile UI Action Sequences
• TutorialVQA: Question Answering Dataset for Tutorial Videos
• Learning to Search in Long Documents Using Document Structure
• Look at the First Sentence: Position Bias in Question Answering
• Improving Human Text Comprehension through Semi-Markov CRF-based Neural Section Title Generation
• Phrase-Indexed Question Answering: A New Challenge for Scalable Document Comprehension
• Multi-Mention Learning for Reading Comprehension with Neural Cascades
• Attentive Convolution: Equipping CNNs with RNN-style Attention Mechanisms
• Using Holographically Compressed Embeddings in Question Answering
• Efficient Conformal Prediction via Cascaded Inference with Expanded Admission
• A Study of the Tasks and Models in Machine Reading Comprehension
• An Empirical Analysis of Multiple-Turn Reasoning Strategies in Reading Comprehension Tasks
• Multi-Perspective Fusion Network for Commonsense Reading Comprehension
• MaP: A Matrix-based Prediction Approach to Improve Span Extraction in Machine Reading Comprehension
• Is Sluice Resolution really just Question Answering?
• TREND: Trigger-Enhanced Relation-Extraction Network for Dialogues
• CAESAR: Context Awareness Enabled Summary-Attentive Reader