Fast and optimal decoding for machine translation

A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most likely according to a set of previously learned parameters (and a formula for combining them). Since the space of possible translations i...

Full description

Saved in:
Bibliographic Details
Published in:Artificial intelligence Vol. 154; no. 1; pp. 127 - 143
Main Authors: Germann, Ulrich, Jahr, Michael, Knight, Kevin, Marcu, Daniel, Yamada, Kenji
Format: Journal Article
Language:English
Published: Elsevier B.V 01.04.2004
Subjects:
ISSN:0004-3702, 1872-7921
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most likely according to a set of previously learned parameters (and a formula for combining them). Since the space of possible translations is extremely large, typical decoding algorithms are only able to examine a portion of it, thus risking to miss good solutions. Unfortunately, examining more of the space leads to unacceptably slow decodings. In this paper, we compare the speed and output quality of a traditional stack-based decoding algorithm with two new decoders: a fast but non-optimal greedy decoder and a slow but optimal decoder that treats decoding as an integer-programming optimization problem.
AbstractList A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most likely according to a set of previously learned parameters (and a formula for combining them). Since the space of possible translations is extremely large, typical decoding algorithms are only able to examine a portion of it, thus risking to miss good solutions. Unfortunately, examining more of the space leads to unacceptably slow decodings. In this paper, we compare the speed and output quality of a traditional stack-based decoding algorithm with two new decoders: a fast but non-optimal greedy decoder and a slow but optimal decoder that treats decoding as an integer-programming optimization problem.
A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most likely according to a set of previously learned parameters (and a formula for combining them). Since the space of possible translations is extremely large, typical decoding algorithms are only able to examine a portion of it, thus risking to miss good solutions. Unfortunately, examining more of the space leads to unacceptably slow decodings. In this paper, we compare the speed and output quality of a traditional stack-based decoding algorithm with two new decoders: a fast but non-optimal greedy decoder and a slow but optimal decoder that treats decoding as an integer-programming optimization problem. (Original abstract)
A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most likely according to a set of previously learned parameters (and a formula for combining them). Since the space of possible translations is extremely large, typical decoding algorithms are only able to examine a portion of it, thus risking to miss good solutions. Unfortunately, examining more of the space leads to unacceptably slow decodings. In this paper, we compare the speed and output quality of a traditional stack-based decoding algorithm with two new decoders: a fast but non-optimal greedy decoder and a slow but optimal decoder that treats decoding as an integer-programming optimization problem.
Author Germann, Ulrich
Knight, Kevin
Yamada, Kenji
Jahr, Michael
Marcu, Daniel
Author_xml – sequence: 1
  givenname: Ulrich
  surname: Germann
  fullname: Germann, Ulrich
  email: germann@isi.edu
  organization: Information Sciences Institute and Department of Computer Science, University of Southern California, 4676 Admiralty Way, Suite 1001, Marina del Rey, CA 90292, USA
– sequence: 2
  givenname: Michael
  surname: Jahr
  fullname: Jahr, Michael
  email: jahr@cs.stanford.edu
  organization: Department of Computer Science, Stanford University, Stanford, CA 94305, USA
– sequence: 3
  givenname: Kevin
  surname: Knight
  fullname: Knight, Kevin
  email: knight@isi.edu
  organization: Information Sciences Institute and Department of Computer Science, University of Southern California, 4676 Admiralty Way, Suite 1001, Marina del Rey, CA 90292, USA
– sequence: 4
  givenname: Daniel
  surname: Marcu
  fullname: Marcu, Daniel
  email: marcu@isi.edu
  organization: Information Sciences Institute and Department of Computer Science, University of Southern California, 4676 Admiralty Way, Suite 1001, Marina del Rey, CA 90292, USA
– sequence: 5
  givenname: Kenji
  surname: Yamada
  fullname: Yamada, Kenji
  email: kyamada@isi.edu
  organization: Information Sciences Institute and Department of Computer Science, University of Southern California, 4676 Admiralty Way, Suite 1001, Marina del Rey, CA 90292, USA
BookMark eNqFkMFKAzEQhoNUsK2-gYc9ees6SXY3Gw-CFKtCwYueQ8zOaso2qUkq-PamricPeplh4P9-mG9GJs47JOScQkmBNpebUodkXSoZAC-hKQHoEZnSVrCFkIxOyBQAqgUXwE7ILMZNPrmUdEroSsdUaNcVfpfsVg9Fh8Z31r0WvQ_FVps367BIQbs46GS9OyXHvR4inv3sOXle3T4t7xfrx7uH5c16YXjTpDxBvrC2rVtegWDYMCEE1YBa87aXTd0bKoWQBgwAQg0aW96LhtedFLpmfE4uxt5d8O97jEltbTQ4DNqh30dVi1pWVMC_QdZSyQFkDl6NQRN8jAF7ZWz6_im_ZwdFQR10qo0adaqDTgWNyjozXP2CdyH7Cp__YdcjhlnVh8WgorHoDHY2oEmq8_bvgi-6qZGN
CitedBy_id crossref_primary_10_1016_j_engappai_2023_106977
crossref_primary_10_1109_ACCESS_2018_2883738
crossref_primary_10_1007_s10590_012_9134_0
crossref_primary_10_1109_ACCESS_2019_2911656
crossref_primary_10_1016_j_datak_2018_05_007
crossref_primary_10_1016_j_procs_2020_03_306
crossref_primary_10_1080_08839510701492546
crossref_primary_10_1111_j_1944_9720_2006_tb02276_x
crossref_primary_10_1016_j_cogsys_2007_10_002
crossref_primary_10_1145_3627168
crossref_primary_10_1145_1380584_1380586
Cites_doi 10.1038/22055
10.1147/rd.136.0675
10.1109/TSSC.1968.300136
ContentType Journal Article
Copyright 2003 Elsevier B.V.
Copyright_xml – notice: 2003 Elsevier B.V.
DBID 6I.
AAFTH
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
E3H
F2A
DOI 10.1016/j.artint.2003.06.001
DatabaseName ScienceDirect Open Access Titles
Elsevier:ScienceDirect:Open Access
CrossRef
Computer and Information Systems Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
Library & Information Sciences Abstracts (LISA)
Library & Information Science Abstracts (LISA)
DatabaseTitle CrossRef
Computer and Information Systems Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Advanced Technologies Database with Aerospace
ProQuest Computer Science Collection
Computer and Information Systems Abstracts Professional
Library and Information Science Abstracts (LISA)
DatabaseTitleList
Library and Information Science Abstracts (LISA)
Computer and Information Systems Abstracts
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1872-7921
EndPage 143
ExternalDocumentID 10_1016_j_artint_2003_06_001
S0004370203001723
GroupedDBID --K
--M
--Z
-~X
.DC
.~1
0R~
1B1
1~.
1~5
23N
4.4
457
4G.
5GY
5VS
6I.
6J9
6TJ
7-5
71M
77K
8P~
9JN
AACTN
AAEDT
AAEDW
AAFTH
AAIAV
AAIKJ
AAKOC
AAKPC
AALRI
AAOAW
AAQFI
AAQXK
AAXUO
AAYFN
ABBOA
ABFNM
ABFRF
ABJNI
ABMAC
ABVKL
ABXDB
ABYKQ
ACDAQ
ACGFO
ACGFS
ACNCT
ACNNM
ACRLP
ACWUS
ACZNC
ADBBV
ADEZE
ADMUD
AEBSH
AECPX
AEFWE
AEKER
AENEX
AETEA
AEXQZ
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
ASPBG
AVWKF
AXJTR
AZFZN
BJAXD
BKOJK
BLXMC
CS3
E3Z
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
F0J
F5P
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
G8K
GBLVA
GBOLZ
HLZ
HVGLF
HZ~
IHE
IXB
J1W
JJJVA
KOM
KQ8
LG9
LY7
M41
MO0
MVM
N9A
NCXOZ
O-L
O9-
OAUVE
OK1
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
R2-
RIG
RNS
ROL
RPZ
SBC
SDF
SDG
SDP
SES
SET
SEW
SPC
SPCBC
SST
SSV
SSZ
T5K
TAE
TN5
TR2
TWZ
UPT
UQL
VQA
WH7
WUQ
XFK
XJE
XJT
XPP
XSW
ZMT
~02
~G-
77I
9DU
AATTM
AAXKI
AAYWO
AAYXX
ABDPE
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
ADVLN
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
7SC
8FD
JQ2
L7M
L~C
L~D
E3H
F2A
ID FETCH-LOGICAL-c366t-c309b2885834072e627771a0eaa38f965fc19779c0c00e050ae83f7635d97a523
ISICitedReferencesCount 20
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000220064400004&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0004-3702
IngestDate Sat Sep 27 22:28:30 EDT 2025
Sun Sep 28 11:26:24 EDT 2025
Sat Nov 29 05:29:00 EST 2025
Tue Nov 18 22:28:06 EST 2025
Fri Feb 23 02:27:13 EST 2024
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 1
Keywords Decoding
SMT
Statistical machine translation
Machine translation
MT
Language English
License http://www.elsevier.com/open-access/userlicense/1.0
https://www.elsevier.com/tdm/userlicense/1.0
https://www.elsevier.com/open-access/userlicense/1.0
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c366t-c309b2885834072e627771a0eaa38f965fc19779c0c00e050ae83f7635d97a523
Notes ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ObjectType-Article-1
ObjectType-Feature-2
OpenAccessLink https://dx.doi.org/10.1016/j.artint.2003.06.001
PQID 28193009
PQPubID 23500
PageCount 17
ParticipantIDs proquest_miscellaneous_57594170
proquest_miscellaneous_28193009
crossref_citationtrail_10_1016_j_artint_2003_06_001
crossref_primary_10_1016_j_artint_2003_06_001
elsevier_sciencedirect_doi_10_1016_j_artint_2003_06_001
PublicationCentury 2000
PublicationDate 2004-04-01
PublicationDateYYYYMMDD 2004-04-01
PublicationDate_xml – month: 04
  year: 2004
  text: 2004-04-01
  day: 01
PublicationDecade 2000
PublicationTitle Artificial intelligence
PublicationYear 2004
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Hart, Nilsson, Raphael (BIB005) 1968; 4
Selman, Levesque, Mitchell (BIB011) 1992
Tillmann, Vogel, Ney, Zubiaga (BIB012) 1997
P. Brown, J. Cocke, S. Della Pietra, V. Della Pietra, F. Jelinek, J. Lai, R. Mercer, Method and system for natural language translation, US Patent 5,477,451, 1995
Knight (BIB007) 1999; 25
Monasson, Zecchina, Kirkpatrick, Selman, Troyansky (BIB008) 1999; 800
Garey, Johnson (BIB004) 1979
Wang, Waibel (BIB013) 1997
Och, Ueffing, Ney (BIB009) 2001
Wu (BIB014) 1996
Papineni, Roukos, Ward, Henderson, Reeder (BIB010) 2002
Brown, Della Pietra, Della Pietra, Mercer (BIB003) 1993; 19
Jelinek (BIB006) 1969; 13
Brown, Cocke, Della Pietra, Della Pietra, Jelinek, Lafferty, Mercer, Roossin (BIB001) 1990; 16
Brown (10.1016/j.artint.2003.06.001_BIB003) 1993; 19
Knight (10.1016/j.artint.2003.06.001_BIB007) 1999; 25
10.1016/j.artint.2003.06.001_BIB002
Wang (10.1016/j.artint.2003.06.001_BIB013) 1997
Wu (10.1016/j.artint.2003.06.001_BIB014) 1996
Hart (10.1016/j.artint.2003.06.001_BIB005) 1968; 4
Jelinek (10.1016/j.artint.2003.06.001_BIB006) 1969; 13
Garey (10.1016/j.artint.2003.06.001_BIB004) 1979
Och (10.1016/j.artint.2003.06.001_BIB009) 2001
Brown (10.1016/j.artint.2003.06.001_BIB001) 1990; 16
Selman (10.1016/j.artint.2003.06.001_BIB011) 1992
Tillmann (10.1016/j.artint.2003.06.001_BIB012) 1997
Papineni (10.1016/j.artint.2003.06.001_BIB010) 2002
Monasson (10.1016/j.artint.2003.06.001_BIB008) 1999; 800
References_xml – start-page: 124
  year: 2002
  end-page: 127
  ident: BIB010
  article-title: Corpus-based comprehensive and diagnostic MT evaluation: Initial Arabic, Chinese, French, and Spanish results
  publication-title: Proceedings of the Human Language Technology Conference, San Diego, CA
– volume: 16
  start-page: 79
  year: 1990
  end-page: 85
  ident: BIB001
  article-title: A statistical approach to machine translation
  publication-title: Comput. Linguistics
– start-page: 440
  year: 1992
  end-page: 446
  ident: BIB011
  article-title: A new method for solving hard satisfiability problems
  publication-title: Proceedings of AAAI-92, San Jose, CA
– start-page: 152
  year: 1996
  end-page: 158
  ident: BIB014
  article-title: A polynomial-time algorithm for statistical machine translation
  publication-title: Proceedings of the 34th ACL
– volume: 13
  start-page: 675
  year: 1969
  end-page: 685
  ident: BIB006
  article-title: A fast sequential decoding algorithm using a stack
  publication-title: IBM Res. J. Res. Development
– reference: P. Brown, J. Cocke, S. Della Pietra, V. Della Pietra, F. Jelinek, J. Lai, R. Mercer, Method and system for natural language translation, US Patent 5,477,451, 1995
– year: 1979
  ident: BIB004
  article-title: Computers and Intractability. A Guide to the Theory of NP-Completeness
– volume: 19
  start-page: 263
  year: 1993
  end-page: 311
  ident: BIB003
  article-title: The mathematics of statistical machine translation: Parameter estimation
  publication-title: Computat. Linguistics
– start-page: 289
  year: 1997
  end-page: 296
  ident: BIB012
  article-title: A DP-based search using monotone alignments in statistical translation
  publication-title: Proceedings of the 35th ACL
– start-page: 55
  year: 2001
  end-page: 62
  ident: BIB009
  article-title: An efficient A
  publication-title: Proceedings of the ACL Workshop on Data-Driven Machine Translation, Toulouse, France
– start-page: 366
  year: 1997
  end-page: 372
  ident: BIB013
  article-title: Decoding algorithm in statistical machine translation
  publication-title: Proceedings of the 35th ACL
– volume: 25
  start-page: 607
  year: 1999
  end-page: 615
  ident: BIB007
  article-title: Decoding complexity in word-replacement translation models
  publication-title: Comput. Linguistics
– volume: 800
  start-page: 133
  year: 1999
  end-page: 137
  ident: BIB008
  article-title: Determining computational complexity from characteristic ‘phrase transitions’
  publication-title: Nature
– volume: 4
  start-page: 100
  year: 1968
  end-page: 107
  ident: BIB005
  article-title: A formal basis for the heuristic determination of minimum cost paths
  publication-title: IEEE Trans. System Sci. Cybernet.
– start-page: 289
  year: 1997
  ident: 10.1016/j.artint.2003.06.001_BIB012
  article-title: A DP-based search using monotone alignments in statistical translation
– start-page: 440
  year: 1992
  ident: 10.1016/j.artint.2003.06.001_BIB011
  article-title: A new method for solving hard satisfiability problems
– start-page: 366
  year: 1997
  ident: 10.1016/j.artint.2003.06.001_BIB013
  article-title: Decoding algorithm in statistical machine translation
– volume: 19
  start-page: 263
  issue: 2
  year: 1993
  ident: 10.1016/j.artint.2003.06.001_BIB003
  article-title: The mathematics of statistical machine translation: Parameter estimation
  publication-title: Computat. Linguistics
– start-page: 55
  year: 2001
  ident: 10.1016/j.artint.2003.06.001_BIB009
  article-title: An efficient A∗ search algorithm for statistical machine translation
– volume: 800
  start-page: 133
  issue: 8
  year: 1999
  ident: 10.1016/j.artint.2003.06.001_BIB008
  article-title: Determining computational complexity from characteristic ‘phrase transitions’
  publication-title: Nature
  doi: 10.1038/22055
– start-page: 152
  year: 1996
  ident: 10.1016/j.artint.2003.06.001_BIB014
  article-title: A polynomial-time algorithm for statistical machine translation
– volume: 13
  start-page: 675
  year: 1969
  ident: 10.1016/j.artint.2003.06.001_BIB006
  article-title: A fast sequential decoding algorithm using a stack
  publication-title: IBM Res. J. Res. Development
  doi: 10.1147/rd.136.0675
– volume: 25
  start-page: 607
  issue: 4
  year: 1999
  ident: 10.1016/j.artint.2003.06.001_BIB007
  article-title: Decoding complexity in word-replacement translation models
  publication-title: Comput. Linguistics
– year: 1979
  ident: 10.1016/j.artint.2003.06.001_BIB004
– start-page: 124
  year: 2002
  ident: 10.1016/j.artint.2003.06.001_BIB010
  article-title: Corpus-based comprehensive and diagnostic MT evaluation: Initial Arabic, Chinese, French, and Spanish results
– volume: 16
  start-page: 79
  issue: 2
  year: 1990
  ident: 10.1016/j.artint.2003.06.001_BIB001
  article-title: A statistical approach to machine translation
  publication-title: Comput. Linguistics
– ident: 10.1016/j.artint.2003.06.001_BIB002
– volume: 4
  start-page: 100
  issue: 2
  year: 1968
  ident: 10.1016/j.artint.2003.06.001_BIB005
  article-title: A formal basis for the heuristic determination of minimum cost paths
  publication-title: IEEE Trans. System Sci. Cybernet.
  doi: 10.1109/TSSC.1968.300136
SSID ssj0003991
Score 1.8983089
Snippet A good decoding algorithm is critical to the success of any statistical machine translation system. The decoder's job is to find the translation that is most...
SourceID proquest
crossref
elsevier
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 127
SubjectTerms Artificial intelligence
Decoding
Machine translation
SMT
Statistical machine translation
Title Fast and optimal decoding for machine translation
URI https://dx.doi.org/10.1016/j.artint.2003.06.001
https://www.proquest.com/docview/28193009
https://www.proquest.com/docview/57594170
Volume 154
WOSCitedRecordID wos000220064400004&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1872-7921
  dateEnd: 20180131
  omitProxy: false
  ssIdentifier: ssj0003991
  issn: 0004-3702
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lj9MwELZglwMX3ojl6QPilpUdN34cV6gVoFI4tKg3y0kcwWo3LU2L9uczfqUtq1XhwMWKojwUf854PPN5PoTe0rrijHFYm-SNyAaKlPDP0SpjclDngqqa-mDOt7GYTOR8rr5GifvOywmItpVXV2r5X6GGcwC22zr7D3D3D4UTcAygQwuwQ_tXwI9MF1jjC7AGly4HAyvMOhEmLz150jppiLa72MLSF6L13CGv5LFTrLNn6Tg7HjzeWc-9Md9Xe2HVUJxkL4T6Gfp4s-UXpyjDLjklWc4BGCOybzlD_ec0ROI-xmAJadjyHydVGmoxXbPXIXRwfupLJnhqKzsN6aHt_JRy8pMvejQbj_V0OJ--W_7MnHKYy7BHGZXb6DgXhQLjfHz2cTj_1M_H4IJF3cTwBWkDpWf5XX_xTQ7KH1O19z-mD9C9uHDAZwHwh-iWbR-h-0mUA0cb_RhRhz8G_HHEHyf8MeCPI_54B_8naDYaTt9_yKIuRlYxztfQElXmUhaSufJ2ludCCGqINYbJRvGiqSi49aoiFSGWFMRYyRpXebBWwhQ5e4qO2kVrnyEsSMkK7rYvlw041twUkrtOLEuYfIpcniCWOkNXsWi80y650IkdeK5DFzo9U6YDSfIEZf1dy1A05cD1IvWzjo5fcOg0jJMDd75JsGiwiy7ZZVq72HTaJYgZLCBuvsJJ0w6oIM8PPuMFurv9J16io_VqY1-hO9Wv9Y9u9TqOt98pwIsR
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Fast+and+optimal+decoding+for+machine+translation&rft.jtitle=Artificial+intelligence&rft.au=Germann%2C+U&rft.au=Jahr%2C+M&rft.au=Knight%2C+K&rft.au=Marcu%2C+D&rft.date=2004-04-01&rft.issn=0004-3702&rft.volume=154&rft.issue=1-2&rft.spage=127&rft.epage=143&rft_id=info:doi/10.1016%2Fj.artint.2003.06.001&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0004-3702&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0004-3702&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0004-3702&client=summon