Addressing Troublesome Words in Neural Machine Translation | |
Zhao, Yang1; Zhang, Jiajun1; He, Zhongjun2; Zong, Chengqing1; Wu, Hua2 | |
2018 | |
会议日期 | 2018-11 |
会议地点 | Brussels, Belgium |
英文摘要 | One of the weaknesses of Neural Machine Translation (NMT) is in handling lowfrequency and ambiguous words, which we refer as troublesome words. To address this problem, we propose a novel memoryenhanced NMT method. First, we investigate different strategies to define and detect the troublesome words. Then, a contextual memory is constructed to memorize which target words should be produced in what situations. Finally, we design a hybrid model to dynamically access the contextual memory so as to correctly translate the troublesome words. The extensive experiments on Chineseto-English and English-to-German translation tasks demonstrate that our method significantly outperforms the strong baseline models in translation quality, especially in handling troublesome words. |
语种 | 英语 |
内容类型 | 会议论文 |
源URL | [http://ir.ia.ac.cn/handle/173211/23196] |
专题 | 自动化研究所_模式识别国家重点实验室_自然语言处理团队 |
作者单位 | 1.中国科学院自动化研究所 2.百度 |
推荐引用方式 GB/T 7714 | Zhao, Yang,Zhang, Jiajun,He, Zhongjun,et al. Addressing Troublesome Words in Neural Machine Translation[C]. 见:. Brussels, Belgium. 2018-11. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
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。
修改评论