| Status | 已發表Published |
| The UM-NLP2CT Neural Machine Translation System for CWMT2017 Translation Task | |
Xu, M. ; Li, Q.; Ao, C. H.; Li, Y.; Chao, S.; Wong, F. 
| |
| 2017-09-01 | |
| Source Publication | The 13th China Workshop on Machine Translation
 |
| Pages | 1-13 |
| Abstract | This paper describes the submission of the University of Macau (UM) for English-to-Chinese news translation task to the 13th China Workshop of Machine Translation (CWMT2017). Our neural machine translation (NMT) system is based on NiuTrans, an attentional encoder-decoder model, extended with several techniques to boost the translation performance. For translation with limited vocabulary, the words referring to time, date and number are respectively generalized as class symbols and translated in the post-processing. In addition, the target vocabularies are further compressed by segmenting the rare words into sub-word units. All the translation systems are trained on the parallel data together with back-translated monolingual data provided by CWMT2017. Our primary system is ranked second in this translation task. |
| Keyword | Neural Machine Translation NLP2CT English-Chinese |
| Language | 英語English |
| The Source to Article | PB_Publication |
| PUB ID | 31250 |
| Document Type | Conference paper |
| Collection | DEPARTMENT OF COMPUTER AND INFORMATION SCIENCE |
| Corresponding Author | Wong, F. |
| Recommended Citation GB/T 7714 | Xu, M.,Li, Q.,Ao, C. H.,et al. The UM-NLP2CT Neural Machine Translation System for CWMT2017 Translation Task[C], 2017, 1-13. |
| APA | Xu, M.., Li, Q.., Ao, C. H.., Li, Y.., Chao, S.., & Wong, F. (2017). The UM-NLP2CT Neural Machine Translation System for CWMT2017 Translation Task. The 13th China Workshop on Machine Translation, 1-13. |
| Files in This Item: | There are no files associated with this item. | |||||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment