본 논문은 영어 문서를 대상으로 문맥의존 철자오류 문제를 해결하고자 한다. 철자오류 종류는 두 가지로 단순 철자오류와 문맥의존 철자오류로 나뉜다. 단순 철자오류는 사전의 단어와 매...

http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
https://www.riss.kr/link?id=T15938543
부산 : 부산대학교 대학원, 2021
학위논문(박사) -- 부산대학교 대학원 , 정보융합공학과 컴퓨터공학전공 , 2021. 8
2021
한국어
006.32 판사항(23)
부산
viii, 86 장 : 삽화, 표 ; 30 cm
지도교수: 권혁철
참고문헌: 장 78-83
I804:21016-000000151875
0
상세조회0
다운로드본 논문은 영어 문서를 대상으로 문맥의존 철자오류 문제를 해결하고자 한다. 철자오류 종류는 두 가지로 단순 철자오류와 문맥의존 철자오류로 나뉜다. 단순 철자오류는 사전의 단어와 매...
본 논문은 영어 문서를 대상으로 문맥의존 철자오류 문제를 해결하고자 한다. 철자오류 종류는 두 가지로 단순 철자오류와 문맥의존 철자오류로 나뉜다. 단순 철자오류는 사전의 단어와 매칭만으로 오류를 찾을 수 있기 때문에 교정하기가 쉽지만 문맥의존 철자오류는 교정 대상 단어와 주변 문맥의 관계를 파악해야 오류 유무를 알 수 있기 때문에 교정의 난이도가 높아진다. 문맥오류의 세부 종류로 동음이의어 오류(homophone error), 문자 배열의 오류(typographical error), 문법 오류(grammatical error), 띄어쓰기 오류(cross word boundary error)로 나뉘며, 논문에서는 띄어쓰기 오류를 제외한 문맥의존 철자오류에 해당하는 나머지 오류에 대해서 다룬다. 그리고 문맥의존 철자오류의 검색은 통계적 방식을 사용하며, 최종 교정어 선택은 딥러닝(deep learning) 방식을 사용하여 문맥의존 철자오류 문제를 해결한다. 논문에서는 기존 문맥의존 철자오류 교정에 다뤄지지 않은 여러 뉴럴 언어모형을 교정에 적용 한다. 논문에서 제안하는 뉴럴 언어모형을 이용한 교정 기법은 크게 5가지로 Word embedding 정보 기반의 교정, Contextual embedding 정보 기반의 교정, Auto-regressive(AR) 계열 언어모형 기반의 교정, Auto-encoding(AE) 계열 언어모형 기반의 교정, Encoder- Decoder 계열 언어모형 기반의 교정으로 나뉜다. 본 논문에서는 최근까지 발표된 15가지 뉴럴 언어모형을 이용해서 문맥의존 철자오류 교정 실험을 진행한다. 논문에서는 교정 대상 단어를 기준으로 양방향의 문맥 정보를 참조하여 교정을 실험하며, 단방향으로 들어오는 입력이나 파라미터 조절을 이용한 성능 실험도 진행하였다. 성능의 측정은 오류어 검색(detection), 오류어 교정(correction)을 각각 정확도(precistion), 재현율(recall), F1으로 표현한다. 논문에서는 문맥의존 철자오류 교정 테스트 말뭉치 구축에 관한 내용도 다루며, 웹에서 얻어진 1조 어절로 구성된 말뭉치를 이용해 실제 사용자들의 오류를 추출하여 성능 테스트에서 제시한다.
목차 (Table of Contents)
참고문헌 (Reference)
1. Deep learning, Goodfellow , Ian, Vol . 1 . No . 2 . Cambridge : MIT press, , 20162016
2. Liu 2019 ] Liu, Peter J.Yu-An ChungJie Ren, `` SummAE : Zero-shot abstractive text summarization using length-agnostic auto-encoders . '' arXiv preprint arXiv:1910.00998, , 2019
3. Peters 2018 ] Peters, Matthew E., `` Deep contextualized word representations . '' arXiv preprint arXiv:1802.05365, , 2018
4. `` Brown corpus manual, FrancisW. NelsonHenry Kucera, '' Letters to the Editor 5.2, , 19791979
5. `` Feed-forward neural networks, Bebis , GeorgeMichael Georgiopoulos, '' IEEE Potentials 13.4, , 19941994
6. `` Context based spelling correction, Mays , EricFred J. DamerauRobert L. Mercer, '' Information Processing & Management 27.5, , 19911991
7. `` Language models are few-shot learners, Brown , Tom B., '' arXiv preprint arXiv:2005.14165, , 20202020
8. `` Neural network acceptability judgments, Warstadt , AlexAmanpreet SinghSamuel R. Bowman, '' Transactions of the Association for Computational Linguistics 7, , 20192019
9. J ? rgen Schmidhuber`` Long short-term memory, Hochreiter , Sepp, '' Neural computation 9.8, , 19971997
10. `` Probability scoring for spelling correction, Kenneth W.William A. Gale, '' Statistics and Computing 1.2, , 19911991
1. Deep learning, Goodfellow , Ian, Vol . 1 . No . 2 . Cambridge : MIT press, , 20162016
2. Liu 2019 ] Liu, Peter J.Yu-An ChungJie Ren, `` SummAE : Zero-shot abstractive text summarization using length-agnostic auto-encoders . '' arXiv preprint arXiv:1910.00998, , 2019
3. Peters 2018 ] Peters, Matthew E., `` Deep contextualized word representations . '' arXiv preprint arXiv:1802.05365, , 2018
4. `` Brown corpus manual, FrancisW. NelsonHenry Kucera, '' Letters to the Editor 5.2, , 19791979
5. `` Feed-forward neural networks, Bebis , GeorgeMichael Georgiopoulos, '' IEEE Potentials 13.4, , 19941994
6. `` Context based spelling correction, Mays , EricFred J. DamerauRobert L. Mercer, '' Information Processing & Management 27.5, , 19911991
7. `` Language models are few-shot learners, Brown , Tom B., '' arXiv preprint arXiv:2005.14165, , 20202020
8. `` Neural network acceptability judgments, Warstadt , AlexAmanpreet SinghSamuel R. Bowman, '' Transactions of the Association for Computational Linguistics 7, , 20192019
9. J ? rgen Schmidhuber`` Long short-term memory, Hochreiter , Sepp, '' Neural computation 9.8, , 19971997
10. `` Probability scoring for spelling correction, Kenneth W.William A. Gale, '' Statistics and Computing 1.2, , 19911991
11. `` Binary codes capable of correcting deletions, LevenshteinVladimir I, insertions , and reversals . '' Soviet physics doklady . Vol . 10 . No . 8 . 1966, , 1966
12. `` Glove : Global vectors for word representation, Pennington , JeffreyRichard SocherChristopher Manning, '' Proceedings of the 2014 conference on empirical methods in natural language processing ( EMNLP ), , 20142014
13. `` NeuSpell : A Neural Spelling Correction Toolkit, Sai MuralidharDanish PruthiGraham Neubig, '' arXiv preprint arXiv:2010.11085, , 20202020
14. `` OCR error correction using a noisy channel model, Kolak , OkanPhilip Resnik, '' Human Language Technology Conference ( HLT 2002, , 20022002
15. `` Categorising spelling errors to assess L2 writing, Bestgen , YvesSylviane Granger, '' International Journal of Continuing Engineering Education and Life-Long Learning 21.2-3, , 20112011
16. `` Language models are unsupervised multitask learners, Radford , Alec, '' OpenAI blog 1.8, , 20192019
17. `` An overview of multi-task learning in deep neural networks, Ruder , Sebastian, '' arXiv preprint arXiv:1706.05098, , 20172017
18. “ A maximum entropy approach to natural language processing, Berger , Adam L.Vincent J. Della PietraStephen A. Della Pietra, '' Computational linguistics , vol.22 , no.1 , page 39-71, , 19961996
19. `` Neural machine translation of rare words with subword units, Sennrich , RicoBarry HaddowAlexandra Birch, '' arXiv preprint arXiv:1508.07909, , 20152015
20. `` Efficient estimation of word representations in vector space, Mikolov , Tomas, '' arXiv preprint arXiv : 1301.3781, , 20132013
21. `` Discriminative syntactic language modeling for speech recognition, Collins , MichaelBrian RoarkMurat Saraclar, '' Proceedings of the 43rd Annual Meeting on Association for Computational Linguistics . Association for Computational Linguistics , page 507-514, , 20052005
22. `` The noisy channel model for unsupervised word sense disambiguation, Yuret , DenizMehmet Ali Yatbaz, '' Computational Linguistics 36.1, , 20102010
23. `` Neural machine translation by jointly learning to align and translate, Bahdanau , DzmitryKyunghyun ChoYoshua Bengio, '' arXiv preprint arXiv:1409.0473, , 20142014
24. `` Spanbert : Improving pre-training by representing and predicting spans, Joshi , Mandar, '' Transactions of the Association for Computational Linguistics 8 ( 2020 ) : 64-77, , 2020
25. `` Neural networks for text correction and completion in keyboard decoding, Ghosh , ShaonaPer Ola Kristensson, '' arXiv preprint arXiv:1709.06429, , 20172017
26. `` Xlnet : Generalized autoregressive pretraining for language understanding, Yang , Zhilin, '' Advances in neural information processing systems, , 20192019
27. `` Parsing natural scenes and natural language with recursive neural networks, Socher , Richard ,, '' ICML, , 20112011
28. `` Distributed representations of words and phrases and their compositionality, Mikolov , Tomas, '' Advances in neural information processing systems, , 20132013
29. `` Exploring the limits of transfer learning with a unified text-to-text transformer, Raffel , Colin, '' arXiv preprint arXiv:1910.10683, , 20192019
30. Li 2018`` Spelling Error Correction Using a Nested RNN Model and Pseudo Training Data, Li , Hao, '' arXiv preprint arXiv:1811.00238, , 2018
31. `` Glue : A multi-task benchmark and analysis platform for natural language understanding, Wang , Alex, '' arXiv preprint arXiv:1804.07461, , 20182018
32. J.H Lee 2018`` Context -Sensitive Spelling Error Correction Techniques using Word Embedding, Jung-Hun LeeMinho KimHyuk-Chul Kwon, '' Proc . of the KIISE Korea Computer Congress 2018 , pp . 607-609, , 2018
33. M.H Kim 2014 ] Minho Kim`` Context -sensitive Spelling Error Correction using Eojeol N-gram, Hyuk-Chul KwonSungki Choi, '' Jounal of KIISE , 41.12 , 1081-1089, , 2014
34. `` An analysis of ill-formed input in natural language queries to document retrieval systems, Charlene W.Caroline M. EastmanRobert L. Oakman, '' Information Processing & Management 27.6, , 19911991
35. J.H Lee`` Context -Sensitive Spelling Error Correction Techniques using Contextual Embeddings, Jung-Hun LeeMinho KimHyuk-Chul Kwon, '' Proc . of the KIISE Korea Computer Congress 2019 , pp . 497-499, , 20192019
36. J.H Lee`` Improved Statistical Language Model for Context-sensitive Spelling Error Candidates, Jung-Hun LeeMinho KimHyuk-Chul Kwon, '' Jounal of Korea Multimedia Society , 20.2 , 371-381, , 20172017
37. `` Google 's neural machine translation system : Bridging the gap between human and machine translation, Wu , Yonghui, '' arXiv preprint arXiv:1609.08144, , 20162016
38. [Radford 2018] Radford, Alec, et al. "Improving language understanding by generative pre-training." (2018)., Radford , Alec, `` Improving language understanding by generative pre-training . '', , 20182018
39. [Vaswani 2017] Vaswani, Ashish, et al. "Attention is all you need." arXiv preprint arXiv:1706.03762 (2017).1, Vaswani , Ashish, `` Attention is all you need'' arXiv preprint arXiv:1706.03762, , 20172017
40. [Dai 2015] Dai, Andrew M., and Quoc V. Le. "Semi-supervised sequence learning." arXiv preprint arXiv:1511.01432 (2015)., DaiAndrew M.Quoc V. Le, `` Semi-supervised sequence learning . '' arXiv preprint arXiv:1511.01432, , 20152015
41. [Edunov 2018] Edunov, Sergey, et al. "Understanding back-translation at scale." arXiv preprint arXiv:1808.09381 (2018)., Edunov , Sergey, `` Understanding back-translation at scale'' arXiv preprint arXiv:1808.09381, , 20182018
42. [Zaheer 2020] Zaheer, Manzil, et al. "Big bird: Transformers for longer sequences." arXiv preprint arXiv:2007.14062 (2020)., Zaheer , Manzil, `` Big bird : Transformers for longer sequences . '' arXiv preprint arXiv:2007.14062, , 20202020
43. [Zhang 2020] Zhang, Shaohua, et al. "Spelling error correction with softmasked BERT." arXiv preprint arXiv:2005.07421 (2020)., Zhang , Shaohua, `` Spelling error correction with softmasked BERT'' arXiv preprint arXiv:2005.07421, , 20202020
44. [Hermann 2015] Hermann, Karl Moritz, et al. "Teaching machines to read and comprehend." arXiv preprint arXiv:1506.03340 (2015)., Karl Moritz, `` Teaching machines to read and comprehend . '' arXiv preprint arXiv:1506.03340, , 20152015
45. [Liu 2019] Liu, Yinhan, et al. "Roberta: A robustly optimized bert pretraining approach." arXiv preprint arXiv:1907.11692 (2019)., Liu , Yinhan, `` Roberta : A robustly optimized bert pretraining approach . '' arXiv preprint arXiv:1907.11692, , 20192019
46. [Bojanowski 2016] Bojanowski, Piotr, et al. "Enriching word vectors with subword information." arXiv preprint arXiv:1607.04606, 2016., Bojanowski , Piotr, `` Enriching word vectors with subword information'' arXiv preprint arXiv:1607.04606, , 20162016
47. [Wing 1980] Wing, Alan M. "Spelling errors in handwriting: A corpus and distributional analysis." Cognitive processes in spelling (1980)., Wing, '' Cognitive processes in spelling, , 19801980
48. `` Adaptive Context-Sensitive Spelling Error Correction Techniques for The Extremely Unpredictable Error Generating Language Environments, Jingzhi JinSungki ChoiHyuk-Chul Kwon, '' Proc . of the KIISE Korea Computer Congress 2015 , pp . 654-656, , 2015Jun . 2015
49. [Conneau 2019] Conneau, Alexis, et al. "Unsupervised cross-lingual representation learning at scale." arXiv preprint arXiv:1911.02116 (2019)., Conneau , Alexis, `` Unsupervised cross-lingual representation learning at scale'' arXiv preprint arXiv:1911.02116, , 20192019
50. [Liu 2019] Liu, Xiaodong, et al. "Multi-task deep neural networks for natural language understanding." arXiv preprint arXiv:1901.11504 (2019)., Liu , Xiaodong, `` Multi-task deep neural networks for natural language understanding'' arXiv preprint arXiv:1901.11504, , 20192019
51. [Rajpurkar 2016] Rajpurkar, Pranav, et al. "Squad: 100,000+ questions for machine comprehension of text." arXiv preprint arXiv:1606.05250 (2016)., Rajpurkar , Pranav, `` Squad : 100,000+ questions for machine comprehension of text'' arXiv preprint arXiv:1606.05250, , 20162016
52. [Beltagy 2020] Beltagy, Iz, Matthew E. Peters, and Arman Cohan. "Longformer: The long-document transformer." arXiv preprint arXiv:2004.05150 (2020)., Beltagy , IzMatthew E. PetersArman Cohan, `` Longformer : The long-document transformer . '' arXiv preprint arXiv:2004.05150, , 20202020
53. [Wiseman 2016] Wiseman, Sam, and Alexander M. Rush. "Sequence-to-sequence learning as beam-search optimization." arXiv preprint arXiv:1606.02960 (2016)., Wiseman, `` Sequence-to-sequence learning as beam-search optimization'' arXiv preprint arXiv:1606.02960, , 20162016
54. [Wang 2019] Wang, Alex, et al. "Superglue: A stickier benchmark for general-purpose language understanding systems." arXiv preprint arXiv:1905.00537 (2019)., Wang , Alex, `` Superglue : A stickier benchmark for general-purpose language understanding systems'' arXiv preprint arXiv:1905.00537, , 20192019
55. [Devlin 2018] Devlin, Jacob, et al. "Bert: Pre-training of deep bidirectional transformers for language understanding." arXiv preprint arXiv:1810.04805 (2018)., Jacob ,, `` Bert : Pre-training of deep bidirectional transformers for language understanding'' arXiv preprint arXiv:1810.04805, , 20182018
56. [He 2016] He, Kaiming, et al. "Deep residual learning for image recognition." Proceedings of the IEEE conference on computer vision and pattern recognition. 2016., He , Kaiming, `` Deep residual learning for image recognition . '' Proceedings of the IEEE conference on computer vision and pattern recognition, , 20162016
57. [Lewis 2019] Lewis, Mike, et al. "Bart: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension." arXiv preprint arXiv:1910.13461 (2019)., Lewis , Mike, `` Bart : Denoising sequence-to-sequence pre-training for natural language generation , translation , and comprehension . '' arXiv preprint arXiv:1910.13461, , 20192019
58. [Zhu 2015] Zhu, Yukun, et al. "Aligning books and movies: Towards story-like visual explana-tions by watching movies and reading books." Proceedings of the IEEE international conference on computer vision. 2015., Zhu , Yukun, `` Aligning books and movies : Towards story-like visual explana-tions by watching movies and reading books . '' Proceedings of the IEEE international conference on computer vision, , 20152015