UM  > 科技學院
Exploiting Rich Feature Representation for SMT N-Best Reranking
Tong, Yu; Derek F. Wong; Lidia S. Chao
2016
Conference NameInternational Conference on Wavelet Analysis and Pattern Recognition (ICWAPR)
Source PublicationPROCEEDINGS OF 2016 INTERNATIONAL CONFERENCE ON WAVELET ANALYSIS AND PATTERN RECOGNITION (ICWAPR)
Pages101-106
Conference DateJUL 10-13, 2016
Conference PlaceJeju, SOUTH KOREA
Abstract

N-best reranking in statistical machine translation (SMT) aims to rescore the possible translation hypotheses such that the best hypothesis appears on the top of the list. The current models either use the standard features (a.k.a model features) that are derived from the SMT decoder or the manually crafted features for reranking. However, in practice, the results are disappointing which had only very small effect on the overall performance. In this paper, we investigate using additional semantic and syntactic feature representations to the reranking framework, with the goal to better capture the differences between best hypothesis and others. These representations are the sentence embeddings that are learned using the recursive autoencoder (RAE). The proposed features are extensively evaluated with various reranking algorithms on WMT2015 French-to-English translation data. The empirical results reveal that the proposed reranking model is able to yield an additional improvement of 1.23 BLEU points over the baseline SMT system.

KeywordN-best Reranking Syntactic Features Semantic Features Deep Neural Networks
DOIhttp://dx.doi.org/10.1109/ICWAPR.2016.7731631
Indexed BySCI
Language英语
WOS Research AreaComputer Science ; Engineering
WOS SubjectComputer Science, Interdisciplinary Applications ; Engineering, Electrical & Electronic
WOS IDWOS:000387487900012
全文获取链接
引用统计
被引频次[WOS]:1   [WOS记录]     [WOS相关记录]
Document TypeConference paper
专题Faculty of Science and Technology
DEPARTMENT OF COMPUTER AND INFORMATION SCIENCE
AffiliationUniv Macau, Dept Comp & Informat Sci, Lab NLP2CT, Macau, Macau Sar, Peoples R China.
First Author AffilicationUniversity of Macau
推荐引用方式
GB/T 7714
Tong, Yu,Derek F. Wong,Lidia S. Chao. Exploiting Rich Feature Representation for SMT N-Best Reranking[C],2016:101-106.
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
Google Scholar
中相似的文章 Google Scholar
[Tong, Yu]的文章
[Derek F. Wong]的文章
[Lidia S. Chao]的文章
Baidu academic
中相似的文章 Baidu academic
[Tong, Yu]的文章
[Derek F. Wong]的文章
[Lidia S. Chao]的文章
Bing Scholar
中相似的文章 Bing Scholar
[Tong, Yu]的文章
[Derek F. Wong]的文章
[Lidia S. Chao]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。