首页> 外文会议>International Conference on Robots Intelligent System >Statistical Machine Translation Algorithm Based on Improved Neural Network
【24h】

Statistical Machine Translation Algorithm Based on Improved Neural Network

机译:基于改进神经网络的统计机器翻译算法

获取原文

摘要

There are rules-based machine translation and modulate-based machine translation but they are all based on complex and hardly-summarizing language rules in essence. This paper discusses necessity and possibility of combination between NN(neural network) and traditional search methods, points advantages and disadvantages of NNMT(neural network machine translation) and puts forward a new MT intelligence integration system framework. It can partially solve some contradictions. If it effectively fuses multi-channel to acquire knowledge such as traditional rules acquisition method, NN method, KDK(knowledge discovery in knowledge base) and KDD(knowledge discovery in database), which largely enhances system solution and relieves bottleneck of grammatical semantic rules acquisition to improve overall performance of machine translation.
机译:基于规则的机器翻译和基于调制的机器翻译,但它们都是基于复杂的和难以概括的语言规则。本文讨论了NN(神经网络)与传统搜索方法之间结合的必要性和可能性,NNMT(神经网络机翻译)的点优缺点,并提出了一种新的MT智能集成系统框架。它可以部分解决一些矛盾。如果有效地保留了多通道以获取传统规则获取方法,NN方法,KDK(知识库中的知识发现)和KDD(数据库中的知识发现)等知识,这在很大程度上增强了系统解决方案并缓解了语法语义规则的瓶颈提高机器翻译的整体性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号