• Publications
  • Influence
DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs
TLDR
We introduce a new reading comprehension benchmark, DROP, which requires Discrete Reasoning Over the content of Paragraphs. Expand
  • 200
  • 42
  • PDF
DuReader: a Chinese Machine Reading Comprehension Dataset from Real-world Applications
  • Wei He, Kai Liu, +10 authors H. Wang
  • Computer Science
  • QA@ACL
  • 14 November 2017
TLDR
This paper introduces DuReader, a new large-scale, open-domain Chinese ma- chine reading comprehension (MRC) dataset, designed to address real-world MRC. Expand
  • 123
  • 30
  • PDF
A Two-Stage Parsing Method for Text-Level Discourse Analysis
TLDR
We propose to use the transition-based model to parse the naked discourse tree (i.e., identifying span and nuclearity) due to data sparsity. Expand
  • 45
  • 19
  • PDF
Toward Fast and Accurate Neural Discourse Segmentation
TLDR
We propose an end-to-end neural discourse segmenter based on BiLSTM-CRF framework and further improve it from two aspects. Expand
  • 30
  • 11
  • PDF
Multi-Passage Machine Reading Comprehension with Cross-Passage Answer Verification
TLDR
In this paper, we propose an end-to-end neural model that enables those answer candidates from different passages to verify each other based on their content representations. Expand
  • 60
  • 10
  • PDF
Do NLP Models Know Numbers? Probing Numeracy in Embeddings
TLDR
We begin by investigating the numerical reasoning capabilities of a state-of-the-art question answering model on the DROP dataset. Expand
  • 60
  • 8
  • PDF
Bag-of-Words as Target for Neural Machine Translation
TLDR
A sentence can be translated into more than one correct sentences. Expand
  • 50
  • 4
  • PDF
Tag-Enhanced Tree-Structured Neural Networks for Implicit Discourse Relation Classification
TLDR
We explore the idea of incorporating syntactic parse tree into neural networks. Expand
  • 10
  • 2
  • PDF
Towards Non-projective High-Order Dependency Parser
TLDR
We use the idea of parsing sequence to bridge the gap between transition-based and graph-based methods under one framework. Expand