期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

基于超图的翻译模型融合的研究

Research on the Translation Model Combination Based on Hypergraph
下载PDF
导出
摘要 当前,系统融合是在机器翻译的后处理上进行.提出了在解码过程中来融合翻译模型,融合了主流两个翻译系统的翻译模型(层次化的基于短语的文法Hiero和括号转录文法BTG).并从理论和实践的角度探索了现在主流的两种解码方法.同时,所提出的解码方法解决了伪歧义或一致性问题.在实验结果上得出:多文法模型融合的标志性要好于成员翻译模型;新的解码方法标志性好于传统解码方法(Viterbi解码). The system combination performs under post-processing, but the paper introduces a translation model combination, which combines two mainstream translation models （Hiero and MaxEnt-based BTG） during decoding. To the spurious ambiguity and consensus problem, the paper introduces new decoding method to solve two problems. In experiment, translation model combination is significantly better than member model, and the new decoding method is better.
作者 刘宇鹏 李生 赵铁军
机构地区 哈尔滨工业大学计算机科学与技术系
出处 《软件学报》 EI CSCD 北大核心 2012年第9期2347-2357,共11页 Journal of Software
基金 国家自然科学基金(60736014) 国家高技术研究发展计划(863)(2006AA010108) 黑龙江省教育厅科学技术研究项目(12521073)
关键词 超图 推导 规则 翻译模型融合 伪歧义 一致性翻译 hypergraph derivation rule translation model combination spurious ambiguity consensus translation
分类号 TP391 [自动化与计算机技术—计算机应用技术]
  • 相关文献

参考文献32

  • 1Rosti AVI, Ayan NF, Xiang B, Matsoukas S, Schwartz R, Dorr B. Combining outputs from multiple machine translation systems. In: Proc. of the HLT/NAACL. New York: Association for Computational Linguistics. 2007. 228-235.
  • 2He XD, Yang M, Gao JF, Nguyen P, Moore R. Indirect-HMMbased hypothesis alignment for combining outputs from machine translation systems. In: Proc. of the EMNLP 2008. Honolulu: Association for Computational Linguistics. 2008.98-107.
  • 3Rosti AV, Matsoukas S, Schwartz R. Improved word-level system combination for machine translation. In: Proc. of the ACL 2007. Prague: Association for Computational Linguistics.2007.312-319.
  • 4Li CH, He XD, Liu YP, Xi N. Incremental HMM alignment for MT system combination. In: Proc. of the ACL 2009. Singapore: Association for Computational Linguistics. 2009. 949-957.
  • 5Chen BX, Zhang M, Li HZ, Aw A. A comparative study of hypothesis alignment and its improvement for machine translation system combination .In: Proc. of the ACL 2009. Singapore: Association for Computational Linguistics, 2009. 941-948.
  • 6Liu YP, Mi H, Feng Y, Liu Q. Joint decoding with multiple translation models. In: Proc. of the ACL 2009. Singapore: Association for Computational Linguistics, 2009. 576-586.
  • 7Li M, Duan N, Zhang D, Li CH, Zhou M. Collaborative decoding: partial hypothesis re-ranking using translation consensus between decoders. In: Proc. of the ACL 2009. Singapore: Association for Computational Linguistics, 2009. 585-592.
  • 8Jiang HF, Yang MY, Zhao TJ, Li S, Wang B. A statistical machine translation model based on a synthetic synchronous grammar. In: Proc. of the ACL 2009. Sin2anore: Association for Comvutational Linguistics. 2009. 125-128.
  • 9Duan N, Li M, Xiao T, Zhou M. The feature subspace method for SMT system combination. In: Proc. of the EMNLP 2009. Honolulu: Association for Computational Linguistics, 2009. 1096-1104.
  • 10Duan N, Li M, Zhang DD, Zhou M. Mixture model-based minimum Bayes risk decoding using multiple machine translation systems. In: Proc. of the COLIN(] 2010. Beijing: Coling 2010 Organizing Committee. 2010. 313-321.
软件学报
2012年 第9期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部