89-95.indd
|
|
- うのすけ あかさか
- 5 years ago
- Views:
Transcription
1 解 説 機械翻訳最新事情 : ( 上 ) 統計的機械翻訳入門 永田昌明渡辺太郎塚田元 NTT 科学基礎研究所 統計的機械翻訳 (statistical machin translation) は, 互いに翻訳になっている 2 つの言語の文の対から翻訳規則や対訳辞書を自動的に学習し, 言語翻訳を実現する技術である. この技術は過去 0 年間に大きく進歩し, アラビア語と英語のような語順が比較的近い言語対では, 従来の翻訳手法より精度が高いと言われている. 本稿では, 上下 2 編に分けて, 近年の自然言語処理で最もホットな話題である統計的機械翻訳の技術概要, および, 評価型ワークショップを中心とした最先端の研究動向を報告する. はじめに Wb (machin translation) 950 (statistical machin translation) Wb 2 統計的機械翻訳とは (knowldg basd machin translation) 2 Googl Translat BETA ATR Trk 情報処理 Vol.49 No. Jan
2 Wb "I would b gratful if you could.." "It would b gratful if you could.." "of " 2 "hip" "hip" "hip joint" 2 2 言語翻訳の生成モデル 980 IB sourc ftargt f 2 "of " "+of" (, f) f P(uf) f P(uf) ê P()P(fu) t = arg max P( ) f = arg max P( ) P ( f ) () () (noisy channl modl) P() ( modl) P(fu) (translation modl) 3 (ncod) P()P(fu) dcod dcodr 2 (paralll corpus) (bilingual corpus) (Hansards) EU (Europarl) 単語に基づく翻訳 990 IB 5 5 IB (word alignmnt) - IB f a 3 () / / 90 情報処理 Vol.49 No. Jan. 2008
3 機械翻訳最新事情 :( 上 ) 統計的機械翻訳入門 is a mans of 単語 is a mans of 句 図 - 単語対応と句対応 P(f, a, ) P(fu) P(f, au) P( f) =! P( fa, ) (2) a IB E IB IB GIZA 8 5 句に基づく翻訳 2000 (phras) "a mans of " "mans" "of" "a" (phras basd T) Kohn 4 f I r I f = rf... rf I r fi r i P(fu) (phras trans lation prob ab ility)f( rfi u r i ) (phras dis tortion probability)d(a i2b i2 ) I p( rf I I r ) = % z ( rfi r i) d( ai-bi-) (3) i= a i i b i2 (i2) count( rf, ) z ( r r f r ) = count( r (4)! fl, r ) fr l count(f, ē) f ē a b, d( a - b ) = a a i - i - - (5) i i- Nagata 6-2 IB (intrsction) (union) (alignmnt point) 情報処理 Vol.49 No. Jan
4 の で ある の で ある 英語から 本語 の単語対応 ( 4) of の で ある is a mans of 本語から英語 の単語対応 ( 4) is a mans of is a mans 言語 言語 は は 本語と英語の対訳句 ( 言語,lana ) ( の,o ) (,comm nication) ( の,a mans o ) ( の,o comm nication) ( の, a mans o comm nication) 対応 図 -2 単語対応付けからの対訳句の抽出 出 : is a mans of is a mans of is a mans of is a mans of is amans of is a mans of is a mans of 出 : である の is a mans of 出 : の である 図 -3 ビーム探索によるデコーディング -3 (mpty) (initial hypothsis) (priority quu) 言語翻訳の識別モデル (gnrativ modl) (discriminativ modl) 2000 (log linar modl) P(uf) h m (, f) 92 情報処理 Vol.49 No. Jan. 2008
5 機械翻訳最新事情 :( 上 ) 統計的機械翻訳入門 (fatur) l m P(uf)!! l! xp m mmhm(, f) = pm ( f ) = (6) xp m h ( l, f) m = m m 6 t = arg max P( f) = arg max! m h (,f) (7) m = h (uf) 5 log p(), h 2 (uf) 5 log p(fu), l 5 l2 5 () l l2 (modl scaling factor) h m (, f) {( s, f s )us 5,... } l (convx) (Gnralizd Itrativ caling) (gradint) t m = arg max! log pm ( s fs) (8) m s = (6) N N bst BLEU(Bilingual Evaluation Undrstudy) 9 BLEU BLEU rfrnc 0 ngram (prcision) p n BP m m BLEU = BP# xp(! log pn) (9) N N n = ngram n ngram p n ngram ngram N54 (8) BLEU Och 7 (inimum Error Rat Training) r E(r, ) BLEU E 5 2BLEU {( s, f s )us 5,...} l t m arg min E(, arg max p =! s m ( fs)) (0) m s = arg max pm ( f s) l f s (0) l l m 構文に基づく翻訳 (syntactic thory) (syntax basd T) 2000 (tr transduction) (Hirarchical Phras Basd Translation) 2 (yn chro nous Contxt Fr Grammar, CFG) 情報処理 Vol.49 No. Jan
6 X <g, a, :> X g a : g a : X X is X () X X X 2, X 2 of X (2) X X X X である X is X X の図 -4 X X of X a mans 階層的句に基づく対訳文の導出 X (3) X (4) X a mans (5) 2 3 (drivation) 2 2 X Chiang 2 2 (glu rul) (chunk) X 2 X 2 (6) X X (7) -4 (initial phras) 2, of, X X of X CKY ngram Watanab 0 Gribach Early Chiang 2 CKY k bst cub pruning Huang Chiang 3 cub pruning (lazy valuation) cub growing おわりに 情報処理 Vol.49 No. Jan. 2008
7 機械翻訳最新事情 :( 上 ) 統計的機械翻訳入門 IB GIZA moss 4 F CPU 参考文献 ) Brown, P. F., Pitra,. A. D., Pitra, V. J. D. and rcr, R. L. : Th athmatics of tatistical achin Translation : Paramtr Estimation, Computational Linguistics, Vol.9, No.2, pp (993). 2) Chiang, D. : Hirarchical Phras Basd Translation, Compu tational Linguistics, Vol.33, No.2, pp (2007). 3) Huang, L. and Chiang, D. : Forst Rscoring : Fastr Dcoding with Intgratd Languag odls, Procdings of th 45th Annual ting of th Association of Computational Linguistics (ACL 07), pp.44 5 (2007). 4) Kohn, P., Och, J. F. and arcu, D. : tatistical Phras Basd Translation, Procdings of th Joint Confrnc on Human Languag Tchnologis and th Annual ting of th North Amrican Chaptr of th Association of Computational Linguistics (HLT NAACL 03), pp (2003). 5) :,, pp.59 28, (2003). 6) Nagata,., aito, K., Yamamoto, K. and Ohashi, K. : A Clustrd Global Phras Rordring odl for tatistical achin Translation, Procdings of th 2st Intrnational Confrnc on Computational Linguistics and 44th Annual ting of th Association for Computational Linguistics (COLING ACL 06), pp (2006). 7) Och, F. J. : inimum Error Rat Training in tatistical achin Translation, Procdings of th 4st Annual ting of th Association for Computational Linguistics (ACL 03), pp (2003). 8) Och, F. J. and Ny, H. : A ystmatic Comparison of Various tatistical Alignmnt odls, Computational Linguistics, Vol.29, No., pp.9 5 (2003). 9) Papinni, K., Roukos,., Ward, T. and Zhu, W. J. : BLEU : a thod for Automatic Evaluation of achin Translation, Procdings of th 40th Annual ting of th Association for Computational Lnguistics (ACL 02), pp.3 38 (2002). 0) Watanab, T., Tsukada, H. and Isozaki, H. : Lft to Right Targt Gnration for Hirarchical Phras Basd Translation, Procdings of th 2st Intrnational Confrnc on Computational Linguistics and 44th Annual ting of th Asso ciation for Computational Linguistics (COLING ACL 06), pp (2006) 永田昌明 ( 正会員 ) nagata.masaaki@lab.ntt.co.jp 987 渡辺太郎 taro@cslab.kcl.ntt.co.jp 2003 塚田元 ( 正会員 ) tsukada@cslab.kcl.ntt.co.jp 989 NTT 4 情報処理 Vol.49 No. Jan
2
NTT 2012 NTT Corporation. All rights reserved. 2 3 4 5 Noisy Channel f : (source), e : (target) ê = argmax e p(e f) = argmax e p(f e)p(e) 6 p( f e) (Brown+ 1990) f1 f2 f3 f4 f5 f6 f7 He is a high school
More informationBLEU Kishore Papineni, Salim Roukos, Todd Ward and Wei-Jing Zhu. (2002) BLEU: a method for Automatic Evaluation of Machine Translation. ACL. MT ( ) MT
4. BLEU @NICT mutiyama@nict.go.jp 1 BLEU Kishore Papineni, Salim Roukos, Todd Ward and Wei-Jing Zhu. (2002) BLEU: a method for Automatic Evaluation of Machine Translation. ACL. MT ( ) MT ( ) BLEU 2 BLEU
More informationIPSJ-TOD
Vol. 3 No. 2 91 101 (June 2010) 1 1 1 2 1 TSC2 Automatic Evaluation of Text Summaries by Using Paraphrase Kazuho Hirahara, 1 Hidetsugu Nanba, 1 Toshiyuki Takezawa 1 and Manabu Okumura 2 The evaluation
More informationA Japanese Word Dependency Corpus ÆüËܸì¤Îñ¸ì·¸¤ê¼õ¤±¥³¡¼¥Ñ¥¹
A Japanese Word Dependency Corpus 2015 3 18 Special thanks to NTT CS, 1 /27 Bunsetsu? What is it? ( ) Cf. CoNLL Multilingual Dependency Parsing [Buchholz+ 2006] (, Penn Treebank [Marcus 93]) 2 /27 1. 2.
More information21 Pitman-Yor Pitman- Yor [7] n -gram W w n-gram G Pitman-Yor P Y (d, θ, G 0 ) (1) G P Y (d, θ, G 0 ) (1) Pitman-Yor d, θ, G 0 d 0 d 1 θ Pitman-Yor G
ol2013-nl-214 No6 1,a) 2,b) n-gram 1 M [1] (TG: Tree ubstitution Grammar) [2], [3] TG TG 1 2 a) ohno@ilabdoshishaacjp b) khatano@maildoshishaacjp [4], [5] [6] 2 Pitman-Yor 3 Pitman-Yor 1 21 Pitman-Yor
More informationIPSJ SIG Technical Report Vol.2014-NL-219 No /12/17 1,a) Graham Neubig 1,b) Sakriani Sakti 1,c) 1,d) 1,e) 1. [23] 1(a) 1(b) [19] n-best [1] 1 N
1,a) Graham Neubig 1,b) Sakriani Sakti 1,c) 1,d) 1,e) 1. [23] 1(a) 1(b) [19] n-best [1] 1 Nara Institute of Science and Technology a) akabe.koichi.zx8@is.naist.jp b) neubig@is.naist.jp c) ssakti@is.naist.jp
More information21 1 2 1 2
21 1 2 1 2 1 2 3 ( ) 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 210 0.0 0.0 22 23 25 27 28 29 30 31 32 33 34 35 36 74 pp.4362003.10 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 141224 14 48 10
More informationNo. 3 Oct The person to the left of the stool carried the traffic-cone towards the trash-can. α α β α α β α α β α Track2 Track3 Track1 Track0 1
ACL2013 TACL 1 ACL2013 Grounded Language Learning from Video Described with Sentences (Yu and Siskind 2013) TACL Transactions of the Association for Computational Linguistics What Makes Writing Great?
More informationtaro.watanabe at nict.go.jp
taro.watanabe at nict.go.jp https://sites.google.com/site/alaginmt2014/ ... I want to study about machine translation. I need to master machine translation. machine translation want to study. infobox infobox
More information数理言語
人工知能特論 II 二宮崇 1 今日の講義の予定 CFG 構文解析 教科書 北研二 ( 著 ) 辻井潤一 ( 編 ) 言語と計算 4 確率的言語モデル東大出版会 C. D. Manning & Hinrich Schütze FOUNDATIONS OF STATISTICAL NATURAL LANGUAGE PROCESSING MIT Press, 1999 D. Jurafsky, J. H.
More information福岡大学人文論叢47-3
679 pp. 1 680 2 681 pp. 3 682 4 683 5 684 pp. 6 685 7 686 8 687 9 688 pp. b 10 689 11 690 12 691 13 692 pp. 14 693 15 694 a b 16 695 a b 17 696 a 18 697 B 19 698 A B B B A B B A A 20 699 pp. 21 700 pp.
More information2
2013 Vol.18 No.2 3 24 25 8 22 2 23 26 9 15 20 2 3 4 5 6 7 8 point1 point 2 point3 point4 10 11 point1 point 2 point 3 point 4 12 13 14 15 16 17 18 19 20 http://www.taishukan.co.jp/kateika/ 21 22 23 24
More information( ) Kevin Duh
NAIST-IS-MT1251045 Factored Translation Models 2014 2 6 ( ) Kevin Duh Factored Translation Models Factored translation models Factored Translation Models, NAIST-IS-MT1251045, 2014 2 6. i Post-ordering
More information,255 7, ,355 4,452 3,420 3,736 8,206 4, , ,992 6, ,646 4,
30 8 IT 28 1,260 3 1 11. 1101. 1102. 1103. 1 3 1,368.3 3 1,109.8 p.5,p.7 2 9,646 4,291 14.5% 10,p.11 3 3,521 8 p.13 45-49 40-44 50-54 019 5 3 1 2,891 3 6 1 3 95 1 1101 1102 1103 1101 1102 1103 1 6,255
More information[1], B0TB2053, 20014 3 31. i
B0TB2053 20014 3 31 [1], B0TB2053, 20014 3 31. i 1 1 2 3 2.1........................ 3 2.2........................... 3 2.3............................. 4 2.3.1..................... 4 2.3.2....................
More informationVol. 23 No. 5 December (Rule-Based Machine Translation; RBMT (Nirenburg 1989)) 2 (Statistical Machine Translation; SMT (Brown, Pietra, Piet
Graham Neubig, Sakriani Sakti 2 Improving Pivot Translation by Remembering the Pivot Akiva Miura, Graham Neubig,, Sakriani Sakti, Tomoki Toda and Satoshi Nakamura In statistical machine translation, the
More informationVol. 23 No. 5 December (Rule-Based Machine Translation; RBMT (Nirenburg 1989)) 2 (Statistical Machine Translation; SMT (Brown, Pietra, Piet
Graham Neubig, Sakriani Sakti 2,,,,, Improving Pivot Translation by Remembering the Pivot Akiva Miura, Graham Neubig,, Sakriani Sakti, Tomoki Toda and Satoshi Nakamura In statistical machine translation,
More informationf ê ê = arg max Pr(e f) (1) e M = arg max λ m h m (e, f) (2) e m=1 h m (e, f) λ m λ m BLEU [11] [12] PBMT 2 [13][14] 2.2 PBMT Hiero[9] Chiang PBMT [X
1,a) Graham Neubig 1,b) Sakriani Sakti 1,c) 1,d) 1,e) 1. Statistical Machine Translation: SMT[1] [2] [3][4][5][6] 2 Cascade Translation [3] Triangulation [7] Phrase-Based Machine Translation: PBMT[8] 1
More informationComputational Semantics 1 category specificity Warrington (1975); Warrington & Shallice (1979, 1984) 2 basic level superiority 3 super-ordinate catego
Computational Semantics 1 category specificity Warrington (1975); Warrington & Shallice (1979, 1984) 2 basic level superiority 3 super-ordinate category preservation 1 / 13 analogy by vector space Figure
More informationNC! 1 1 1/1000
1 2000 NC NC! 1 1 1/1000 1 4 3p. 37. 2 1 2 1 2 NC 2 2 3 1 3 3 3! ! 1 10 10 30! 2 3 1770 20 1950 NC NC 1970 70 82 NC 1970 70 NC NC NC 1 NC NC 4 1p. 97 NC NC 5 NC 1950 NC NC 3pp. 183188 NC! 1 2 3 1 NC 2
More information-----------------------------------------------------------------------------------------1 --------------------------------------------------------------------------------------1 -------------------------------------------------------------------------------------1
More information自然言語処理23_175
2 Sequence Alignment as a Set Partitioning Problem Masaaki Nishino,JunSuzuki, Shunji Umetani, Tsutomu Hirao and Masaaki Nagata Sequence alignment, which involves aligning elements of two given sequences,
More informationRapp BLEU[10] [9] BLEU OrthoBLEU Rapp OrthoBLEU [9] OrthoBLEU OrthoBLEU ) ) ) 1) NTT Natural Language Research
RJ-008 Is Back Translation Really Unuseful? Validation of Back Translation from the Perspective of a Checking Method for Users Mai Miyabe Takashi Yoshino 1. [1, 2] [3] [4] 1 2 2 [3,5,6,7] [8, 9] 1: 2 3
More informationIBM-Mode1 Q: A: cash money It is fine today 2
8. IBM Model-1 @NICT mutiyama@nict.go.jp 1 IBM-Mode1 Q: A: cash money It is fine today 2 e f a P (f, a e) â : â = arg max a P (f, a e) â P (f, a e) 3 θ P (f e, θ) θ f d = { f, e } L(θ d) = log f,e d P
More informationit-ken_open.key
深層学習技術の進展 ImageNet Classification 画像認識 音声認識 自然言語処理 機械翻訳 深層学習技術は これらの分野において 特に圧倒的な強みを見せている Figure (Left) Eight ILSVRC-2010 test Deep images and the cited4: from: ``ImageNet Classification with Networks et
More information[1] B =b 1 b n P (S B) S S O = {o 1,2, o 1,3,, o 1,n, o 2,3,, o i,j,, o n 1,n } D = {d 1, d 2,, d n 1 } S = O, D o i,j 1 i
1,a) 2,b) 3,c) 1,d) CYK 552 1. 2 ( 1 ) ( 2 ) 1 [1] 2 [2] [3] 1 Graduate School of Information Science, Nagoya University, Japan 2 Information Technology Center, Nagoya University, Japan 3 Information &
More information[4], [5] [6] [7] [7], [8] [9] 70 [3] 85 40% [10] Snowdon 50 [5] Kemper [3] 2.2 [11], [12], [13] [14] [15] [16]
1,a) 1 2 1 12 1 2Type Token 2 1 2 1. 2013 25.1% *1 2012 8 2010 II *2 *3 280 2025 323 65 9.3% *4 10 18 64 47.6 1 Center for the Promotion of Interdisciplinary Education and Research, Kyoto University 2
More information名称未設定
NAACL-HLT 2012, 1 2012 6 3 8 NAACL-HLT 2012 (North American Chapter of ACL: Human Language Technologies) ACL Anthology 1 2 NAACL ACL (Association for Computational Linguistics) 2000 2001 2 ACL HLT 2003
More information( )
B4IM2035 2017 2 10 ( ) (e.g., eat ) (e.g., arrest ),,, 10., B4IM2035, 2017 2 i 1 1 2 3 2.1................. 3 2.2........ 3 3 5 3.1.... 5 3.2 DCS Vector.............................. 6 3.3 DCS Vector.......
More informationuntitled
The 8 th Depression Rework Research Association Annual Meeting 8 2015 42526 The 8 th Depression Rework Research Association Annual Meeting 8 2015 42526 INDEX 1 2 3 4 7 10 12 21 27 1 29 2 37 3 41 4 47 5
More information一般社団法人電子情報通信学会 THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGIN
一般社団法人電子情報通信学会 THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS 信学技報 IEICE Technical Report SP2019-12(2019-08)
More information2 3 4 5 6 7 8 9 10 11 Licca is always with you 12 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 30 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31 31
More informationRF_1
RF_1 10/04/16 10:32 http://rftechno.web.infoseek.co.jp/rf_1.html 1/12 RF_1 10/04/16 10:32 http://rftechno.web.infoseek.co.jp/rf_1.html 2/12 RF_1 10/04/16 10:32 http://rftechno.web.infoseek.co.jp/rf_1.html
More informationF8302D_1目次_160527.doc
N D F 830D.. 3. 4. 4. 4.. 4.. 4..3 4..4 4..5 4..6 3 4..7 3 4..8 3 4..9 3 4..0 3 4. 3 4.. 3 4.. 3 4.3 3 4.4 3 5. 3 5. 3 5. 3 5.3 3 5.4 3 5.5 4 6. 4 7. 4 7. 4 7. 4 8. 4 3. 3. 3. 3. 4.3 7.4 0 3. 3 3. 3 3.
More information_4th2..
01-07 章 _4th2 了 18.8.25 6:46 PM ページ 110 110 2060 1569 2059201510 01-07 章 _4th2 了 18.8.25 6:46 PM ページ 111 111 43 26 548 795 254 1 3,686 443 283 1,668 4,129 915 6,712 2710 2730 2710 65 201729 20166,712 3,3834,081
More information!
The 22nd Annual Meeting of the Japanese Association of Cardiac Rehabilitation ! The 22nd Annual Meeting of the Japanese Association of Cardiac Rehabilitation The 22nd Annual Meeting of the Japanese Association
More information卓球の試合への興味度に関する確率論的分析
17 i 1 1 1.1..................................... 1 1.2....................................... 1 1.3..................................... 2 2 5 2.1................................ 5 2.2 (1).........................
More information¥ì¥·¥Ô¤Î¸À¸ì½èÍý¤Î¸½¾õ
2013 8 18 Table of Contents = + 1. 2. 3. 4. 5. etc. 1. ( + + ( )) 2. :,,,,,, (MUC 1 ) 3. 4. (subj: person, i-obj: org. ) 1 Message Understanding Conference ( ) UGC 2 ( ) : : 2 User-Generated Content [
More informationパソコン接続マニュアル P-01F 日本語
P-01F 1 2 3 4 5 1 2 +m1111 1 2 3 4 5 6 6 1 1 111 2 1 3 1 1 1 2 1 7 3 8 1 2 1 111 3 4 5 9 1 m111 m1111 c 2 3 4 5 10 1 1 111 2 1 3 1 1 1 2 1 3 1 m111 m1111 2 3 1 11 12 1 2 3 1 2 3 13 1 2 3 4 5 14 6 7 8 9
More informationohpmain.dvi
fujisawa@ism.ac.jp 1 Contents 1. 2. 3. 4. γ- 2 1. 3 10 5.6, 5.7, 5.4, 5.5, 5.8, 5.5, 5.3, 5.6, 5.4, 5.2. 5.5 5.6 +5.7 +5.4 +5.5 +5.8 +5.5 +5.3 +5.6 +5.4 +5.2 =5.5. 10 outlier 5 5.6, 5.7, 5.4, 5.5, 5.8,
More information2014/1 Vol. J97 D No. 1 2 [2] [3] 1 (a) paper (a) (b) (c) 1 Fig. 1 Issues in coordinating translation services. (b) feast feast feast (c) Kran
a) b) c) Improving Quality of Pivot Translation by Context in Service Coordination Yohei MURAKAMI a), Rie TANAKA b),andtoruishida c) Web 1. Web 26.8% 30.9% 21.3% 21% 1 n n(n 1) Department of Social Informatics,
More informationFR 34 316 13 303 54
FR 34 316 13 303 54 23 ( 1 14 ) ( 3 10 ) 8/4 8/ 100% 8 22 7 12 1 9 8 45 25 28 17 19 14 3/1 6/27 5000 8/4 12/2930 1 66 45 43 35 49 25 22 20 23 21 17 13 20 6 1 8 52 1 50 4 11 49 3/4/5 75 6/7/8 46 9/10/11
More informationMicrosoft PowerPoint - 11Syntax.ppt
言語理論 : 知的情報処理 (11) 構文論 慶應義塾大学理工学部櫻井彰人 言語を定義する方法 ( いくつかある ): 文法 ( 生成規則 ) オートマトン 既知の言語間の演算 これらの間には対応関係がある 自然言語の記述だけでなく 例えば コンパイラの設計に使用 言語の定義方法 プログラム言語の構文を簡単に どうやって定義するか 定義方法は使いやすくあるべし, i..: 定義は有限の長さ 与えられた文字列がその言語に属するか否かを調べるアルゴリズムが存在する必要がある
More information自然言語処理22_289
(RNN) Dyer (Dyer, Clark, Lavie, and Smith 2011) IBM (Brown, Pietra, Pietra, and Mercer 1993) word embedding word embedding RNN (Yang, Liu, Li, Zhou, and Yu 2013) IBM 4 Recurrent Neural Networks for Word
More information坊っちゃん
might is right I am glad to see you 1992 4 1 20 1 2 1987 62 10 27 1 5-86 1999
More information*p145-174_Œâ‡í‡ê‡é
*p145-174_ 問 われる 09.1.16 10:34 PM ページ145 2007 200708 146 147 a s 148 a s d f g 153 a s d 158 a s d f g h j 166 a s d f 171 2009 145 *p145-174_ 問 われる 09.1.16 10:34 PM ページ146 45 2007 2008 146 *p145-174_
More information2 of 46 07.2.10 4:30 PM
1 of 46 07.2.10 4:30 PM 2 of 46 07.2.10 4:30 PM 3 of 46 07.2.10 4:30 PM 4 of 46 07.2.10 4:30 PM 5 of 46 07.2.10 4:30 PM 6 of 46 07.2.10 4:30 PM 7 of 46 07.2.10 4:30 PM 8 of 46 07.2.10 4:30 PM 9 of 46 07.2.10
More informationP1-1 P1-2 P1-3 P1-4 P1-5 P1-6 P3-1 P3-2 P3-3 P3-4 P3-5 P3-6 P5-1 P5-2 P5-3 P5-4 P5-5 P5-6 P7-1 P7-2 P7-3 P7-4 P7-5 P7-6 P9-1 P9-2 P9-3 P9-4 P9-5 P9-6 P11-1 P11-2 P11-3 P11-4 P13-1 P13-2 P13-3 P13-4 P13-5
More informationgengo.dvi
4 97.52% tri-gram 92.76% 98.49% : Japanese word segmentation by Adaboost using the decision list as the weak learner Hiroyuki Shinnou In this paper, we propose the new method of Japanese word segmentation
More informationc 2009 i
I 2009 c 2009 i 0 1 0.0................................... 1 0.1.............................. 3 0.2.............................. 5 1 7 1.1................................. 7 1.2..............................
More information統計的機械翻訳モデルの構築 各モデルを対訳文から学習 対訳文 太郎が花子を訪問した Taro visited Hanako. 花子にプレセントを渡した He gave Hanako a present.... モデル翻訳モデル並べ替えモデル言語モデル 2
ALAGIN 機械翻訳セミナー単語アライメント Graham Neubig 奈良先端科学技術大学院大学 (NAIST) 2014 年 3 月 5 日 https://sites.google.com/site/alaginmt2014/ 1 統計的機械翻訳モデルの構築 各モデルを対訳文から学習 対訳文 太郎が花子を訪問した Taro visited Hanako. 花子にプレセントを渡した He gave
More informationフレーズベース機械翻訳システムの構築 フレーズベース機械翻訳システムの構築 Graham Neubig & Kevin Duh 奈良先端科学技術大学院大学 (NAIST) 5/10/2012 1
Graham Neubig & Kevin Duh 奈良先端科学技術大学院大学 (NAIST) 5/10/2012 1 フレーズベース統計的機械翻訳 ( SMT ) 文を翻訳可能な小さい塊に分けて 並べ替える Today I will give a lecture on machine translation. Today 今日は I will give を行います a lecture on の講義
More information野岩鉄道の旅
29th 5:13 5:34 5:56 6:00 6:12 6:20 6:21 6:25 6:29 6:31 6:34 6:38 6:40 6:45 6:52 6:56 7:01 7:07 7:11 7:32 7:34 7:50 7:58 8:03 8:17 8:36 8:44 5:50 5:54 6:15 6:38 6:39 6:51 6:59 6:59 7:03 7:08 7:08 7:11 7:15
More informationJRSJ Vol. 23 No. 8 2 Nov., 2005
23 8 1 2005 11 JRSJ Vol. 23 No. 8 2 Nov., 2005 23 8 3 2005 11 JRSJ Vol. 23 No. 8 4 Nov., 2005 23 8 5 2005 11 JRSJ Vol. 23 No. 8 6 Nov., 2005 23 8 7 2005 11 JRSJ Vol. 23 No. 8 8 Nov., 2005 23 8 9 2005 11
More informationIPSJ SIG Technical Report Vol.2010-NL-199 No /11/ treebank ( ) KWIC /MeCab / Morphological and Dependency Structure Annotated Corp
1. 1 1 1 2 treebank ( ) KWIC /MeCab / Morphological and Dependency Structure Annotated Corpus Management Tool: ChaKi Yuji Matsumoto, 1 Masayuki Asahara, 1 Masakazu Iwatate 1 and Toshio Morita 2 This paper
More informationuntitled
10 23 11 1 1 (3) 40 0 2,861 2,697 571 6.041 75,628 6,062 229 442 2 3 4 5 6 7 8 9 10 11 12 JIS B 856230 13 14 15 16 17 18 19 20 21 22 23 24 1 25 NTT au NTT i NTT au NTT 26 27 161 (17,102 )10 23 25 (FM)
More informationuntitled
580 26 5 SP-G 2011 AI An Automatic Question Generation Method for a Local Councilor Search System Yasutomo KIMURA Hideyuki SHIBUKI Keiichi TAKAMARU Hokuto Ototake Tetsuro KOBAYASHI Tatsunori MORI Otaru
More information1 Twitter Twitter Twitter 2. 1 Xu [3] Twitter Twitter Twitter Twitter iphone iphone iphone Twitter Xu [3] Twitter Xu [5] Web Web Web Web
DEIM Forum 2015 G8-5 Twitter 305 8550 1 2 305 8550 1 2 E-mail: s1111509@u.tsukuba.ac.jp, yohei@slis.tsukuba.ac.jp n-gram (n=3 9) 18 Twitter 1. 1. 1 Twitter 1 Twitter Twitter Twitter Twitter 1https://twitter.com
More informationmain.dvi
Vol. 44 No. 11 Nov. 2003 2 (1) (2) Exploring Transfer Errors in Lexical and Structural Paraphrasing Atsushi Fujita and Kentaro Inui In lexical and structural paraphrasing, meaning-preserving linguistic
More information3 4 26 1980 1 WWW 26! 3, ii 4 7!! 4 2010 8 1. 1.1... 1 1.2... 2 1.3... 3 1.4... 7 1.5... 9... 9 2. 2.1... 10 2.2... 13 2.3... 16 2.4... 18... 21 3. 3.1... 22 3.2... 24 3.3... 33... 38 iv 4. 4.1... 39 4.2...
More information日経テレコン料金表(2016年4月)
1 2 3 4 8,000 15,000 22,000 29,000 5 6 7 8 36,000 42,000 48,000 54,000 9 10 20 30 60,000 66,000 126,000 166,000 50 100 246,000 396,000 1 25 8,000 7,000 620 2150 6,000 4,000 51100 101200 3,000 1,000 201
More information122011pp.139174 18501933
122011pp.139174 18501933 122011 1850 3 187912 3 1850 8 1933 84 4 1871 12 1879 5 2 1 9 15 1 1 5 3 3 3 6 19 9 9 6 28 7 7 4 1140 9 4 3 5750 58 4 3 1 57 2 122011 3 4 134,500,000 4,020,000 11,600,000 5 2 678.00m
More informationMicrosoft Word - 映画『東京裁判』を観て.doc
1 2 3 4 5 6 7 1 2008. 2 2010, 3 2010. p.1 4 2008 p.202 5 2008. p.228 6 2011. 7 / 2008. pp.3-4 1 8 1 9 10 11 8 2008, p.7 9 2011. p.41 10.51 11 2009. p. 2 12 13 14 12 2008. p.4 13 2008, p.7-8 14 2008. p.126
More information() L () 20 1
() 25 1 10 1 0 0 0 1 2 3 4 5 6 2 3 4 9308510 4432193 L () 20 1 PP 200,000 P13P14 3 0123456 12345 1234561 2 4 5 6 25 1 10 7 1 8 10 / L 10 9 10 11 () ( ) TEL 23 12 7 38 13 14 15 16 17 18 L 19 20 1000123456
More information308 ( ) p.121
307 1944 1 1920 1995 2 3 4 5 308 ( ) p.121 309 10 12 310 6 7 ( ) ( ) ( ) 50 311 p.120 p.142 ( ) ( ) p.117 p.124 p.118 312 8 p.125 313 p.121 p.122 p.126 p.128 p.156 p.119 p.122 314 p.153 9 315 p.142 p.153
More information73 p.1 22 16 2004p.152
1987 p.80 72 73 p.1 22 16 2004p.152 281895 1930 1931 12 28 1930 10 27 12 134 74 75 10 27 47.6 1910 1925 10 10 76 10 11 12 139 p.287 p.10 11 pp.3-4 1917 p.284 77 78 10 13 10 p.6 1936 79 15 15 30 80 pp.499-501
More information