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
1NTTTDDCITMja-en2019/07/26 13:35:373002---57.34------SMTNoTransformer Big Single Fine-tuned with items data R2L re-ranking 4 models ensemble
2NTTTDDCITMja-en2019/07/23 13:03:382851---56.93------NMTNoTransformer Big Single Fine-tuned with items data 4 models ensemble
3NTTTDDCITMja-en2019/07/22 16:20:562810---56.14------NMTNoTransformer Big Single Fine-tuned with items data
4NTTTDDCITMja-en2019/07/22 14:40:222801---54.58------NMTNoTransformer Big Single
5geoduckTDDCITMja-en2019/07/27 13:47:383197---54.27------OtherYesTM + NMT (a single Transformer)
6sarahTDDCITMja-en2019/07/22 16:18:162807---54.25------NMTNoTransformer, ensemble of 4 models
7NICT-2TDDCITMja-en2019/07/26 22:25:513081---53.87------NMTNoTransformer, ensemble of 4 models w/ long warm-up and self-training
8sarahTDDCITMja-en2019/07/22 16:23:162811---52.77------NMTNoTransformer, single model
9NICT-2TDDCITMja-en2019/07/26 22:24:083080---52.55------NMTNoTransformer, sigle model w/ long warm-up and self-training
10geoduckTDDCITMja-en2019/07/27 16:36:463215---50.90------OtherYesTM + MT (ensemble)
11geoduckTDDCITMja-en2019/07/27 16:42:073216---46.21------OtherYesTM + MT (ensemble)
12ORGANIZERTDDCITMja-en2019/08/02 16:45:273262---38.55------NMTNobaseline using OpenNMT
13ORGANIZERTDDCITMja-en2019/08/02 16:51:163264---33.75------OtherYesAzure Custom Translator

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
1NTTTDDCITMja-en2019/07/26 13:35:373002---0.844086------SMTNoTransformer Big Single Fine-tuned with items data R2L re-ranking 4 models ensemble
2NTTTDDCITMja-en2019/07/23 13:03:382851---0.842011------NMTNoTransformer Big Single Fine-tuned with items data 4 models ensemble
3NTTTDDCITMja-en2019/07/22 16:20:562810---0.838679------NMTNoTransformer Big Single Fine-tuned with items data
4NICT-2TDDCITMja-en2019/07/26 22:25:513081---0.834898------NMTNoTransformer, ensemble of 4 models w/ long warm-up and self-training
5sarahTDDCITMja-en2019/07/22 16:18:162807---0.832909------NMTNoTransformer, ensemble of 4 models
6NTTTDDCITMja-en2019/07/22 14:40:222801---0.831940------NMTNoTransformer Big Single
7NICT-2TDDCITMja-en2019/07/26 22:24:083080---0.831156------NMTNoTransformer, sigle model w/ long warm-up and self-training
8geoduckTDDCITMja-en2019/07/27 13:47:383197---0.828493------OtherYesTM + NMT (a single Transformer)
9sarahTDDCITMja-en2019/07/22 16:23:162811---0.823336------NMTNoTransformer, single model
10geoduckTDDCITMja-en2019/07/27 16:36:463215---0.781113------OtherYesTM + MT (ensemble)
11ORGANIZERTDDCITMja-en2019/08/02 16:45:273262---0.758265------NMTNobaseline using OpenNMT
12ORGANIZERTDDCITMja-en2019/08/02 16:51:163264---0.733992------OtherYesAzure Custom Translator
13geoduckTDDCITMja-en2019/07/27 16:42:073216---0.703675------OtherYesTM + MT (ensemble)

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
1NTTTDDCITMja-en2019/07/26 13:35:373002---0.816660------SMTNoTransformer Big Single Fine-tuned with items data R2L re-ranking 4 models ensemble
2NTTTDDCITMja-en2019/07/23 13:03:382851---0.814690------NMTNoTransformer Big Single Fine-tuned with items data 4 models ensemble
3NTTTDDCITMja-en2019/07/22 16:20:562810---0.808040------NMTNoTransformer Big Single Fine-tuned with items data
4sarahTDDCITMja-en2019/07/22 16:18:162807---0.807250------NMTNoTransformer, ensemble of 4 models
5NICT-2TDDCITMja-en2019/07/26 22:25:513081---0.805460------NMTNoTransformer, ensemble of 4 models w/ long warm-up and self-training
6NTTTDDCITMja-en2019/07/22 14:40:222801---0.804600------NMTNoTransformer Big Single
7geoduckTDDCITMja-en2019/07/27 13:47:383197---0.800360------OtherYesTM + NMT (a single Transformer)
8NICT-2TDDCITMja-en2019/07/26 22:24:083080---0.797870------NMTNoTransformer, sigle model w/ long warm-up and self-training
9sarahTDDCITMja-en2019/07/22 16:23:162811---0.794050------NMTNoTransformer, single model
10geoduckTDDCITMja-en2019/07/27 16:36:463215---0.765690------OtherYesTM + MT (ensemble)
11ORGANIZERTDDCITMja-en2019/08/02 16:51:163264---0.729880------OtherYesAzure Custom Translator
12ORGANIZERTDDCITMja-en2019/08/02 16:45:273262---0.712040------NMTNobaseline using OpenNMT
13geoduckTDDCITMja-en2019/07/27 16:42:073216---0.710630------OtherYesTM + MT (ensemble)

Notice:

Back to top

HUMAN (WAT2019)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description
1sarahTDDCITMja-en2019/07/22 16:18:162807UnderwayNMTNoTransformer, ensemble of 4 models
2sarahTDDCITMja-en2019/07/22 16:23:162811UnderwayNMTNoTransformer, single model
3NTTTDDCITMja-en2019/07/26 13:35:373002UnderwaySMTNoTransformer Big Single Fine-tuned with items data R2L re-ranking 4 models ensemble
4NICT-2TDDCITMja-en2019/07/26 22:25:513081UnderwayNMTNoTransformer, ensemble of 4 models w/ long warm-up and self-training

Notice:
Back to top

HUMAN (WAT2018)


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

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