期刊文献+

翻译评价方法的区分度分析 被引量:1

Discriminability Analysis of Translation Evaluation
下载PDF
导出
摘要 首先简单介绍了目前被广泛应用的两种翻译评价方法BLEU和NIST,然后着眼于它们对翻译系统实际水平的区分程度,参考教育测量领域中的方法,提出了区分度的概念,并通过实验阐明了基于区分度的性能分析方法。实验结果表明这种性能分析方法能够有效地反映出BLEU方法和NIST方法在区分程度上的优劣。 Firstly,this paper simply introduces two widely-applied methods of translation evaluation:BLEU and NIST.Then it focuses on their capability of discriminating translation systems' practical level,referring to the approach in the field of education measure,and proposes a concept of discriminability.An experiment is given to illustrate the performance analysis method based on discriminability.The experimental results indicate this method can effectively reflect the discriminating degree of BLEU and NIST.
出处 《计算机工程与应用》 CSCD 北大核心 2005年第21期94-96,共3页 Computer Engineering and Applications
基金 国家863高技术研究发展计划基金项目(编号:2002AA11701009)
关键词 机器翻译 评价方法 性能分析 区分度 machine translation,evaluating method,performance analysis,discriminability
  • 相关文献

参考文献9

  • 1K Papineni,S Roukos,T Ward et al. BLEU:a method for automatic evaluation of MT[R].IBM research division,T J Watson Research Centre, Research Report: Computer Science RC22176 (W0109-022),2001
  • 2NIST.Automatic evaluation of machine translation quality using Ngram co-occurrence statistics.http://www.nist.gov/speech/tests/mt/doc/ngram-study.pdf, 2003
  • 3Flanagan Mary A.Error Classification for MT Evaluation′ , in Technology Partnerships for Crossing the Language Barrier[C].In:Proceedings of the First Conference of the Association for Machine Translation in the Americas, Columbia,Maryland, 1994: 65~72
  • 4T O′Connell,F O′Mara,J White.The ARPA MT evaluation methodologies: Evolution,lessons and further approaches[C].In: Proceedings of the First Conference of the Association for Machine Translation in the Americas, Columbia, USA, 1994
  • 5H Thompson. The Strategic Role of Evaluation in Natural Language Processing and Speech Technology[R].Technical Report,University of Edinburgh Record of a workshop sponsored by DANDI,ELSNET and HCRC, 1992
  • 6张剑,吴际,周明.机器翻译评测的新进展[J].中文信息学报,2003,17(6):1-8. 被引量:14
  • 7C X Ling,H Zhang. AUC:a statistically consistent and more discriminating measure than accuracy[C].In:International Joint Conference on Artificial Intelligence IJCAI03,2003:519~526
  • 8王孝玲.教育测量[M].上海:华东师范大学出版社,2001.109-230.
  • 9Language, machines: computers in translation and linguistics (ALPAC report)[R].National academy of sciences,Nat′l Academies Press,1966

二级参考文献11

  • 1俞士汶.Automatic Evaluation of Output Quality for Machine Translation Systems .Machine Translation,1993,8:17-126.
  • 2俞士汶 姜新 朱学锋.基于测试集与测试点的机译系统评估[A].见陈肇雄主编.机器翻译研究进展[C].电子工业出版社,1992年.524—537.
  • 3俞士汶 段慧明.英汉机器翻译译文质量测试大纲[J].计算机世界(D版技术专题D10-D11),1998,(13).
  • 4BLEU: a method for automatic evaluation of machine translation [C]. Kishore Papieni, Salim Roukos,Todd Ward, Wei - Jing Zhu, 232 - 240,ACL 2002.
  • 5Hovy, E.H. 1999. Toward Finely Differentiated Evaluation Metrics for Machine Translation [ C]. Proceedings of the EAGLES Workshop on Standards and Evaluation, 439 - 446 Pisa, Italy.
  • 6JEIDA 1992. JEIDA Methodology and Criteria on Machine Translation Evaluation (JEIDA Report)[R]. H. Ncanura (editor). 108- 120, Japan Electronic Industry Development Association.
  • 7King, M. and K. Falkedal. 1990. Using Test Suites in Evaluation of Machine Translation Systems[C]. Proceedings of the 18th (X)LING Conference vol. 2. 18 - 32 Helsinki, Finland.
  • 8Language and Machines. Computers in Translation and Linguistics. (ALPAC report, 1966) , [R]. National Academy of Sciences.
  • 9The EAGLES MT Evaluation Working Group. 1996. EAGLES Evaluation of Natural Language Processing Systems [ R]. Final Report. EAGLES Document EAG- EWG - PR 2, ISBN 87 - 90708 - 00- 8. 550 - 583, Center for Sprogteknologi, Copenhagen.
  • 10Vogel,- S., Niessen S., Hermann Ney. 2000. Automatic Extrapolation of Human Assessment of Translation Quality [ C]. LREC-2000, 487- 499, Athens, Greece.

共引文献27

同被引文献47

引证文献1

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