首页> 外文期刊>Information Technology Journal >Arabic-Chinese and Chinese-Arabic Phrase-Based Statistical Machine Translation Systems
【24h】

Arabic-Chinese and Chinese-Arabic Phrase-Based Statistical Machine Translation Systems

机译:基于阿拉伯语和中文阿拉伯语短语的统计机器翻译系统

获取原文
获取原文并翻译 | 示例
           

摘要

Designs for Arabic-to-Chinese and Chinese-to-Arabic translation systems are presented. The core of the system implements standard Phrase-Based Statistical Machine Translation architecture, where Corpus data used for the systems was collected from the United-Nations website and various news engine websites. Here, we focus on its acquisition as it is the training data of Arabic-Chinese and Chinese-Arabic Statistical Machine Translation systems. We trained Statistical Machine Translation systems for two language pairs, which revealed interesting clues into the challenges ahead. Models are then softly integrated into Statistical Machine Translation architecture so they can interact with other models without modifying the basic architecture. As a result, phrase translation probabilities learn directly rather than deriving them heuristically.
机译:介绍了阿拉伯语到中文和中文到阿拉伯语翻译系统的设计。该系统的核心实现了标准的基于短语的统计机器翻译体系结构,其中用于系统的语料库数据是从联合国网站和各种新闻引擎网站收集的。在这里,我们将重点放在它的获取上,因为它是阿拉伯语-中文和中文-阿拉伯语统计机器翻译系统的训练数据。我们为两种语言对训练了统计机器翻译系统,这为我们面临的挑战提供了有趣的线索。然后将模型软集成到统计机器翻译体系结构中,以便它们可以与其他模型进行交互,而无需修改基本体系结构。结果,短语翻译概率直接学习而不是启发式地推导它们。

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