NICT_LOGO.JPG KYOTO-U_LOGO.JPG

WAT

The Workshop on Asian Translation
Evaluation Results

[EVALUATION RESULTS TOP] | [BLEU] | [RIBES] | [AMFM] | [HUMAN (WAT2019)] | [HUMAN (WAT2018)] | [HUMAN (WAT2017)] | [HUMAN (WAT2016)] | [HUMAN (WAT2015)] | [HUMAN (WAT2014)] | [EVALUATION RESULTS USAGE POLICY]

BLEU


# Team Task Date/Time DataID BLEU
Method
Other
Resources
System
Description
juman kytea mecab moses-
tokenizer
stanford-
segmenter-
ctb
stanford-
segmenter-
pku
indic-
tokenizer
unuse myseg kmseg
1NICT-5INDICta-en2018/08/24 15:00:222133---24.31---- 0.00 0.00NMTNoXX-XX transformer model
2cvitINDICta-en2019/03/23 12:44:442664---23.31---- 0.00 0.00NMTYesmassive-multi e270
3IITP-MTINDICta-en2018/09/14 19:54:082349---22.42---- 0.00 0.00NMTNoTransformer multilingual XX-En
4cvitINDICta-en2019/03/14 22:10:492627---21.94---- 0.00 0.00NMTYesmassive-multi + ft
5NICT-5INDICta-en2018/08/24 14:49:522111---21.37---- 0.00 0.00SMTNoXX-En transformer model
6NICT-5INDICta-en2018/09/07 14:31:492241---21.27---- 0.00 0.00NMTNoUnified Source vocabulary by orthography mapping. XX-EN model.
7cvitINDICta-en2019/03/22 05:44:232655---21.18---- 0.00 0.00NMTYesmay to en (Transformer) - detokenized
8cvitINDICta-en2019/03/22 05:20:192647---20.51---- 0.00 0.00NMTYesmany to en model (Transformer)
9ORGANIZERINDICta-en2018/08/24 14:38:362101---19.71---- 0.00 0.00NMTNomulti2one multilingual NMT with Attention
10ORGANIZERINDICta-en2018/08/29 14:30:012193---18.59---- 0.00 0.00NMTNomulti2multi multilingual NMT with Attention
11cvitINDICta-en2019/03/14 21:54:152620---17.63---- 0.00 0.00NMTYesmassive-multi
12AnuvaadINDICta-en2018/09/15 17:49:302400---14.34---- 0.00 0.00SMTNoSMT with KenLM
13AnuvaadINDICta-en2018/09/15 18:02:352408---14.09---- 0.00 0.00SMTNoSMT XX-En
14RGNLPINDICta-en2018/09/15 02:50:582376---12.96---- 0.00 0.00SMTNoSMT system with SRILM Language model
15RGNLPINDICta-en2018/09/15 02:40:102370---12.90---- 0.00 0.00SMTNoSMT system with KENLM Language model
16RGNLPINDICta-en2018/09/15 03:18:242385---11.70---- 0.00 0.00NMTNoNMT system with a 2-layer LSTM method
17NICT-5INDICta-en2018/08/24 14:49:402110---11.09---- 0.00 0.00NMTNoBilingual transformer model
18ORGANIZERINDICta-en2018/08/20 11:17:412008--- 9.14---- 0.00 0.00NMTNoNMT with Attention

Notice:

Back to top

RIBES


# Team Task Date/Time DataID RIBES
Method
Other
Resources
System
Description
juman kytea mecab moses-
tokenizer
stanford-
segmenter-
ctb
stanford-
segmenter-
pku
indic-
tokenizer
unuse myseg kmseg
1cvitINDICta-en2019/03/23 12:44:442664---0.779394----0.0000000.000000NMTYesmassive-multi e270
2cvitINDICta-en2019/03/14 22:10:492627---0.770894----0.0000000.000000NMTYesmassive-multi + ft
3cvitINDICta-en2019/03/22 05:44:232655---0.769896----0.0000000.000000NMTYesmay to en (Transformer) - detokenized
4NICT-5INDICta-en2018/08/24 15:00:222133---0.768865----0.0000000.000000NMTNoXX-XX transformer model
5IITP-MTINDICta-en2018/09/14 19:54:082349---0.757610----0.0000000.000000NMTNoTransformer multilingual XX-En
6cvitINDICta-en2019/03/22 05:20:192647---0.756278----0.0000000.000000NMTYesmany to en model (Transformer)
7ORGANIZERINDICta-en2018/08/24 14:38:362101---0.751277----0.0000000.000000NMTNomulti2one multilingual NMT with Attention
8ORGANIZERINDICta-en2018/08/29 14:30:012193---0.744884----0.0000000.000000NMTNomulti2multi multilingual NMT with Attention
9NICT-5INDICta-en2018/08/24 14:49:522111---0.744744----0.0000000.000000SMTNoXX-En transformer model
10cvitINDICta-en2019/03/14 21:54:152620---0.742074----0.0000000.000000NMTYesmassive-multi
11NICT-5INDICta-en2018/09/07 14:31:492241---0.740337----0.0000000.000000NMTNoUnified Source vocabulary by orthography mapping. XX-EN model.
12AnuvaadINDICta-en2018/09/15 18:02:352408---0.673058----0.0000000.000000SMTNoSMT XX-En
13AnuvaadINDICta-en2018/09/15 17:49:302400---0.671535----0.0000000.000000SMTNoSMT with KenLM
14RGNLPINDICta-en2018/09/15 03:18:242385---0.670636----0.0000000.000000NMTNoNMT system with a 2-layer LSTM method
15RGNLPINDICta-en2018/09/15 02:50:582376---0.668149----0.0000000.000000SMTNoSMT system with SRILM Language model
16RGNLPINDICta-en2018/09/15 02:40:102370---0.662031----0.0000000.000000SMTNoSMT system with KENLM Language model
17NICT-5INDICta-en2018/08/24 14:49:402110---0.658743----0.0000000.000000NMTNoBilingual transformer model
18ORGANIZERINDICta-en2018/08/20 11:17:412008---0.649417----0.0000000.000000NMTNoNMT with Attention

Notice:

Back to top

AMFM


# Team Task Date/Time DataID AMFM
Method
Other
Resources
System
Description
juman kytea mecab moses-
tokenizer
stanford-
segmenter-
ctb
stanford-
segmenter-
pku
indic-
tokenizer
unuse myseg kmseg
1cvitINDICta-en2019/03/23 12:44:442664---0.613670----0.0000000.000000NMTYesmassive-multi e270
2IITP-MTINDICta-en2018/09/14 19:54:082349---0.604300----0.0000000.000000NMTNoTransformer multilingual XX-En
3cvitINDICta-en2019/03/14 22:10:492627---0.602280----0.0000000.000000NMTYesmassive-multi + ft
4cvitINDICta-en2019/03/22 05:20:192647---0.594690----0.0000000.000000NMTYesmany to en model (Transformer)
5cvitINDICta-en2019/03/22 05:44:232655---0.594690----0.0000000.000000NMTYesmay to en (Transformer) - detokenized
6NICT-5INDICta-en2018/08/24 15:00:222133---0.593410----0.0000000.000000NMTNoXX-XX transformer model
7cvitINDICta-en2019/03/14 21:54:152620---0.578460----0.0000000.000000NMTYesmassive-multi
8ORGANIZERINDICta-en2018/08/24 14:38:362101---0.568020----0.0000000.000000NMTNomulti2one multilingual NMT with Attention
9NICT-5INDICta-en2018/09/07 14:31:492241---0.566780----0.0000000.000000NMTNoUnified Source vocabulary by orthography mapping. XX-EN model.
10ORGANIZERINDICta-en2018/08/29 14:30:012193---0.561550----0.0000000.000000NMTNomulti2multi multilingual NMT with Attention
11NICT-5INDICta-en2018/08/24 14:49:522111---0.552630----0.0000000.000000SMTNoXX-En transformer model
12RGNLPINDICta-en2018/09/15 03:18:242385---0.519100----0.0000000.000000NMTNoNMT system with a 2-layer LSTM method
13RGNLPINDICta-en2018/09/15 02:40:102370---0.517460----0.0000000.000000SMTNoSMT system with KENLM Language model
14AnuvaadINDICta-en2018/09/15 17:49:302400---0.511130----0.0000000.000000SMTNoSMT with KenLM
15RGNLPINDICta-en2018/09/15 02:50:582376---0.510700----0.0000000.000000SMTNoSMT system with SRILM Language model
16ORGANIZERINDICta-en2018/08/20 11:17:412008---0.488060----0.0000000.000000NMTNoNMT with Attention
17AnuvaadINDICta-en2018/09/15 18:02:352408---0.487250----0.0000000.000000SMTNoSMT XX-En
18NICT-5INDICta-en2018/08/24 14:49:402110---0.471260----0.0000000.000000NMTNoBilingual transformer model

Notice:

Back to top

HUMAN (WAT2019)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description

Notice:
Back to top

HUMAN (WAT2018)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description
1IITP-MTINDICta-en2018/09/14 19:54:08234975.250NMTNoTransformer multilingual XX-En
2NICT-5INDICta-en2018/08/24 15:00:22213363.250NMTNoXX-XX transformer model
3NICT-5INDICta-en2018/08/24 14:49:52211147.750SMTNoXX-En transformer model
4AnuvaadINDICta-en2018/09/15 17:49:30240029.750SMTNoSMT with KenLM
5AnuvaadINDICta-en2018/09/15 18:02:35240829.250SMTNoSMT XX-En

Notice:
Back to top

HUMAN (WAT2017)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description

Notice:
Back to top

HUMAN (WAT2016)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description

Notice:
Back to top

HUMAN (WAT2015)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description

Notice:
Back to top

HUMAN (WAT2014)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description

Notice:
Back to top

EVALUATION RESULTS USAGE POLICY

When you use the WAT evaluation results for any purpose such as:
- writing technical papers,
- making presentations about your system,
- advertising your MT system to the customers,
you can use the information about translation directions, scores (including both automatic and human evaluations) and ranks of your system among others. You can also use the scores of the other systems, but you MUST anonymize the other system's names. In addition, you can show the links (URLs) to the WAT evaluation result pages.

NICT (National Institute of Information and Communications Technology)
Kyoto University
Last Modified: 2018-08-02