• Publications
  • Influence
SummaRuNNer: A Recurrent Neural Network Based Sequence Model for Extractive Summarization of Documents
We present SummaRuNNer, a Recurrent Neural Network (RNN) based sequence model for extractive summarization of documents and show that it achieves performance better than or comparable toExpand
Improving the Transformer Translation Model with Document-Level Context
TLDR
This work extends the Transformer model with a new context encoder to represent document-level context, which is then incorporated into the original encoder and decoder, and introduces a two-step training method to take full advantage of abundant sentence-level parallel corpora and limited document- level parallel Corpora. Expand
Neural Models for Sequence Chunking
TLDR
This paper investigates the use of DNN for sequence chunking, and proposes three neural models so that each chunk can be treated as a complete unit for labeling, which can achieve start-of-the-art performance on both the text chunking and slot filling tasks. Expand
Augmenting String-to-Tree Translation Models with Fuzzy Use of Source-side Syntax
TLDR
A new way to use the source syntax in a fuzzy manner, both in source syntactic annotation and in rule matching is proposed, which not only guarantees grammatical output with an explicit target tree, but also enables the system to choose the proper translation rules via fuzzy use of the sourcentax. Expand
Three Strategies to Improve One-to-Many Multilingual Translation
TLDR
This work introduces three strategies to improve one-to-many multilingual translation by balancing the shared and unique features and proposes to divide the hidden cells of the decoder into shared and language-dependent ones. Expand
A Compact and Language-Sensitive Multilingual Translation Method
TLDR
To maximize parameter sharing, this work presents a universal representor to replace both encoder and decoder models, and introduces language-sensitive embedding, attention, and discriminator with the ability to enhance model performance. Expand
Handling Unknown Words in Statistical Machine Translation from a New Perspective
TLDR
A new perspective to handle unknown words in statistical machine translation is proposed, which focuses on determining the semantic function the unknown words serve as in the test sentence and keeping the semanticfunction unchanged in the translation process. Expand
Unsupervised Tree Induction for Tree-based Translation
TLDR
Experimental results show that the string-to-tree translation system using an unsupervised tree structure derived from a novel non-parametric Bayesian model significantly outperforms the strong baseline string- to-tree system using parse trees. Expand
A Substitution-Translation-Restoration Framework for Handling Unknown Words in Statistical Machine Translation
TLDR
Extensive experiments on both phrase-based and linguistically syntax-based SMT models in Chinese-to-English translation show that the new perspective on handling unknown words in statistical machine translation (SMT) can substantially improve the translation quality. Expand
Tree-based Translation without using Parse Trees
TLDR
This paper makes a great effort to bypass the parse trees and induce effective unsupervised trees for treebased translation models and results have shown that the string-to-tree translation system using the unsuper supervised trees significantly outperforms the string to-tree system using parse trees. Expand
...
1
2
...