89-95.indd

Size: px
Start display at page:

Download "89-95.indd"

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

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 information

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 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 information

IPSJ-TOD

IPSJ-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 information

A Japanese Word Dependency Corpus ÆüËܸì¤Îñ¸ì·¸¤ê¼õ¤±¥³¡¼¥Ñ¥¹

A 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 information

21 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

21 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 information

IPSJ 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

IPSJ 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 information

21 1 2 1 2

21 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 information

No. 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

No. 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 information

taro.watanabe at nict.go.jp

taro.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

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

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 information

2

2 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

( ) 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,

,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

[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 information

Vol. 23 No. 5 December (Rule-Based Machine Translation; RBMT (Nirenburg 1989)) 2 (Statistical Machine Translation; SMT (Brown, Pietra, Piet

Vol. 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 information

Vol. 23 No. 5 December (Rule-Based Machine Translation; RBMT (Nirenburg 1989)) 2 (Statistical Machine Translation; SMT (Brown, Pietra, Piet

Vol. 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 information

f 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

f 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 information

Computational 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 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 information

NC! 1 1 1/1000

NC! 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

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 information

Rapp BLEU[10] [9] BLEU OrthoBLEU Rapp OrthoBLEU [9] OrthoBLEU OrthoBLEU ) ) ) 1) NTT Natural Language Research

Rapp 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 information

IBM-Mode1 Q: A: cash money It is fine today 2

IBM-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 information

it-ken_open.key

it-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] 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]

[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

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 information

untitled

untitled 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 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 information

2 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 information

RF_1

RF_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 information

F8302D_1目次_160527.doc

F8302D_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..

_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

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

¥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

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 information

ohpmain.dvi

ohpmain.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 information

2014/1 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

2014/1 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 information

FR 34 316 13 303 54

FR 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 information

Microsoft PowerPoint - 11Syntax.ppt

Microsoft 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

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

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_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 information

2 of 46 07.2.10 4:30 PM

2 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 information

P1-1 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 information

gengo.dvi

gengo.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 information

c 2009 i

c 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

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

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

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 information

JRSJ Vol. 23 No. 8 2 Nov., 2005

JRSJ 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 information

IPSJ SIG Technical Report Vol.2010-NL-199 No /11/ treebank ( ) KWIC /MeCab / Morphological and Dependency Structure Annotated Corp

IPSJ 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 information

untitled

untitled 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 information

untitled

untitled 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 information

1 Twitter Twitter Twitter 2. 1 Xu [3] Twitter Twitter Twitter Twitter iphone iphone iphone Twitter Xu [3] Twitter Xu [5] Web Web Web Web

1 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 information

main.dvi

main.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 information

3 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)

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 information

122011pp.139174 18501933

122011pp.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 information

Microsoft Word - 映画『東京裁判』を観て.doc

Microsoft 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

() 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 information

308 ( ) p.121

308 ( ) 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 information

73 p.1 22 16 2004p.152

73 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