본문 바로가기
KERD
논문KCI 등재

최신 기계번역 품질 예측 연구

Research on Recent Quality Estimation

임희석|문현석|박찬준|서재형|어수경
한국융합학회 (Korea Convergence Society)|2021.07|한국융합학회논문지 (Journal of the Korea Convergence Society)|제12권 제7호|pp. 37-44|DOI https://doi.org/10.15207/JKCS.2021.12.7.037

1

조회

0

다운로드

2

인용

0

참고문헌

키워드

Natural Language ProcessingNeural Machine Translation딥러닝Deep Learning자연언어처리Language Convergence언어 융합인공신경망 기계번역기계번역 품질 예측Quality Estimation
주제분류학제간연구

초록

기계번역 품질 예측(Quality Estimation, QE)은 정답 문장(Reference sentence) 없이도 기계번역 결과의 질을 평가할 수 있으며, 활용도가 높다는 점에서 그 필요성이 대두되고 있다. Conference on machine translation(WMT)에서 매년 이와 관련한 shared task가 열리고 있고 최근에는 대용량 데이터 기반 Pretrained language model(PLM)을 적용한 연구들이 주로 진행되고 있다. 본 논문에서는 기계번역 품질 예측 task에 대한 설명 및 연구 동향에 대한 전반적인 survey를 진행했고, 최근 자주 활용되는 PLM의 특징들에 대해 정리하였다. 더불어 아직 활용된 바가 없는 multilingual BART 모델을 이용하여 기존 연구들인 XLM, multilingual BERT, XLM-RoBERTa와의 비교 실험 및 분석을 진행하였다. 실험 결과 어떤 사전 학습된 다중언어 모델이 QE에 적용했을 때 가장 효과적인지 확인하였을 뿐 아니라 multilingual BART 모델의 QE 태스크 적용 가능성을 확인했다.

Abstract

Quality estimation (QE) can evaluate the quality of machine translation output even for those who do not know the target language, and its high utilization highlights the need for QE. QE shared task is held every year at Conference on Machine Translation (WMT), and recently, researches applying Pretrained Language Model (PLM) are mainly being conducted. In this paper, we conduct a survey on the QE task and research trends, and we summarize the features of PLM. In addition, we used a multilingual BART model that has not yet been utilized and performed comparative analysis with the existing studies such as XLM, multilingual BERT, and XLM-RoBERTa. As a result of the experiment, we confirmed which PLM was most effective when applied to QE, and saw the possibility of applying the multilingual BART model to the QE task.

참고문헌 (35)

  1. L. Specia, "Findings of the wmt 2018 shared task on quality estimation", Association for Computational Linguistics (2018)[DOI]
  2. E. Fonseca, "Findings of the WMT 2019shared tasks on quality estimation", Proceedings of the Fourth Conference on Machine Translation, 3, pp.1~10 (2019)
  3. L. Specia, "QuEst-A translation quality estimation framework", Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pp.79~84 (2013)
  4. Lucia Specia, "Quality Estimation for Machine Translation", Synthesis Lectures on Human Language Technologies, 11(1), pp.1~162 (2018), Morgan & Claypool Publishers LLC[DOI]
  5. Lucia Specia, "Machine translation evaluation versus quality estimation", Machine Translation, 24(1), pp.39~50 (2010), Springer Science and Business Media LLC[DOI]
  6. D. Lee., "Two-phase cross-lingual language model fine-tuning for machine translation quality estimation", Proceedings of the Fifth Conference on Machine Translation, pp.1024~1028 (2020)
  7. Y. Baek, "Patquest: Papago translation quality estimation", Proceedings of the Fifth Conference on Machine Translation, pp.991~998 (2020)
  8. G. Lample, "Cross-lingual language model pretraining", arXiv preprint arXiv:1901.07291., ~ (2019)
  9. J. Devlin, "Bert : Pre-training of deep bidirectional transformers for language understanding", arXiv preprint arXiv:1810.04805, ~ (2018)
  10. A. Conneau, "Unsupervised cross-lingual representation learning at scale", arXiv preprint arXiv:1911.02116, ~ (2019)
  11. Y. Liu, "Multilingual denoising pre-training for neural machine translation", Transactions of the Association for Computational Linguistics, 8, pp.726~742 (2020)
  12. Referential translation machines for predicting translation quality, "E. Biçici.", Association for Computational Linguistics, ~ (2014)
  13. R. Soricut, "The SDL language weaver systems in the WMT12 quality estimation shared task", Proceedings of the Seventh Workshop on Statistical Machine Translation, pp.145~151 (2012)
  14. N. Q. Luong, "LIG system for WMT13 QE task: Investigating the usefulness of features in word confidence estimation for MT", 8th Workshop on Statistical Machine Translation, pp.386~391 (2013)
  15. C. Hardmeier, "Tree kernels for machine translation quality estimation", Seventh Workshop on Statistical Machine Translation, pp.109~113 (2012), Association for Computational Linguistics
  16. R. N. Patel., "Translation quality estimation using recurrent neural network", arXiv preprint arXiv:1610.04841, ~ (2016)
  17. H. Kim, "Recurrent neural network based translation quality estimation", Proceedings of the First Conference on Machine Translation:Volume 2, Shared Task Papers, pp.787~792 (2016)
  18. K. Cho, "Learning phrase representations using RNN encoder-decoder for statistical machine translation", arXiv preprint arXiv:1406.1078, ~ (2014)
  19. Sepp Hochreiter, "Long Short-Term Memory", Neural Computation, 9(8), pp.1735~1780 (1997), MIT Press - Journals[DOI]
  20. H. Kim, "Predictor-estimator using multilevel task learning with stack propagation for neural quality estimation", Proceedings of the Second Conference on Machine Translation, pp.562~568 (2017)
  21. J. Wang, "Alibaba submission for WMT18 quality estimation task", Proceedings of the Third Conference on Machine Translation: Shared Task Papers, pp.809~815 (2018)
  22. A. Vaswani, "Attention is all you need", Advances in neural information processing systems, pp.5998~6008 (2017)
  23. F. Kepler, "Unbabel's Participation in the WMT19 Translation Quality Estimation Shared Task", arXiv preprint arXiv:1907.10352, ~ (2019)
  24. H. Kim, "QE BERT : bilingual BERT using multi-task learning for neural quality estimation", Proceedings of the Fourth Conference on Machine Translation, 3, pp.85~89 (2019)
  25. T. Ranasinghe, "TransQuest at WMT2020 : Sentence-Level Direct Assessment", arXiv preprint arXiv:2010.05318., ~ (2020)
  26. D. Lee., "Two-phase cross-lingual language model fine-tuning for machine translation quality estimation", Proceedings of the Fifth Conference on Machine Translation, pp.1024~1028 (2020)
  27. M. Wang, "Hw-tsc’s participation at wmt 2020 quality estimation shared task", Proceedings of the Fifth Conference on Machine Translation, pp.1056~1061 (2020)
  28. H. Wu, "Tencent submission for WMT20 Quality Estimation Shared Task", Proceedings of the Fifth Conference on Machine Translation, pp.1062~1067 (2020)
  29. M. Snover, "A study of translation edit rate with targeted human annotation", Proceedings of association for machine translation in the Americas, 200(6) (2006)
  30. G. Wenzek, "Ccnet : Extracting high quality monolingual datasets from web crawl data", arXiv preprint arXiv:1911.00359., ~ (2019)
  31. T. Pires, "How multilingual is multilingual bert?", arXiv preprint arXiv:1906.01502, ~ (2019)
  32. M. Lewis, "Bart : Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension", arXiv preprint arXiv:1910.13461, ~ (2019)
  33. T. Wolf, "HuggingFace's Transformers : State-of-the-art natural language processing", arXiv preprint arXiv:1910.03771, ~ (2019)
  34. 박찬준, "공공 한영 병렬 말뭉치를 이용한 기계번역 성능 향상 연구", 디지털융복합연구, 18(6), pp.271~277 (2020), 한국디지털정책학회[DOI]
  35. C. Park, "Decoding strategies for improving low-resource machine translation", Electronics, 9(10), pp.1562~ (2020)

관련 논문AI추천

Zero-Shot 기반 기계번역 품질 예측 연구문현석, 박찬준, 서재형, 어수경, 임희석한국융합학회 · 한국융합학회논문지2021 · 제12권 제11호병렬 말뭉치 필터링을 적용한 Filter-mBART기반 기계번역 연구문현석, 박정배, 박찬준, 어수경, 임희석한국융합학회 · 한국융합학회논문지2021 · 제12권 제5호신경망기계번역의 객관적 평가를 위한 예비연구: 자동평가와 수동평가의 균형점이준호통번역연구소 · 통번역학연구2019 · 제23권 제3호버트(BERT)를 활용한 인간번역의 자동평가: 여러 모델의 성능 비교 및 활용 가능성서수영, 정혜연통번역연구소 · 통번역학연구2022 · 제26권 제4호신경망기계번역 기술 진화와 번역품질 분석김희동, 지인영통번역연구소 · 통번역학연구2020 · 제24권 제3호통역 자동 평가의 가능성과 한계 고찰- MTQE 적용 사례를 중심으로-최문선통번역연구소 · 통번역학연구2025 · 제29권 제1호

이 간행물의 다른 논문 (34건)

방송기능이 있는 IP PBX 융합 중앙 관제 시스템 개발김삼택pp. 1-6온라인 쇼핑의 데이터 융합 기반 사이즈 추천 서비스: 서비스 품질, 정보 신뢰, 고객 만족의 구매 의도에 대한 역할김지은pp. 7-17게임을 읽는 머리, 스포츠지능을 측정하기 위한 컴퓨터 기반 스포츠 인지검사 프로그램 개발박진한, 우민정pp. 19-29Research model on stock price prediction system through real-time Macroeconomics index and stock news mining analysis홍성혁pp. 31-36빅데이터 수집 처리를 위한 분산 하둡 풀스택 플랫폼의 설계이명호pp. 45-51
원문 보기