NICT_LOGO.JPG KYOTO-U_LOGO.JPG

WAT

The Workshop on Asian Translation
Evaluation Results

[EVALUATION RESULTS TOP] | [BLEU] | [RIBES] | [AMFM] | [HUMAN (WAT2022)] | [HUMAN (WAT2021)] | [HUMAN (WAT2020)] | [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
1HW-TSCINDIC20ml-en2020/09/19 11:47:564042---20.87------NMTNotransformer deep(pre LN),xx2en,PMI data and filtered PBI data (fasttext domain adaptation)+ sentencepiece + 2 model ensemble
2cvitINDIC20ml-en2020/09/18 00:22:373843---16.98------NMTYes
3cvitINDIC20ml-en2020/09/18 00:35:003849---15.93------NMTYes
4ORGANIZERINDIC20ml-en2020/09/02 16:44:303634---15.46------NMTNoBaseline MLNMT XX to En model using PIB and Filtered PMI data. Transformer big model. Default settings.
5cvitINDIC20ml-en2020/10/14 13:00:324146---15.33------NMTNoBaseline NMT fine-tuned then pre-trained on filtered Cyclic BT corpus further fine-tune.
6ODIANLPINDIC20ml-en2020/09/17 21:41:123821---15.30------NMTNoTransformer Big with Relative position representations + xx-en model + PMI Data
7NICT-5INDIC20ml-en2020/09/18 21:22:264004---13.50------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size unbalanced.
8NICT-5INDIC20ml-en2020/09/18 21:01:373991---11.44------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size balanced.

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
1cvitINDIC20ml-en2020/09/18 00:22:373843---0.724107------NMTYes
2cvitINDIC20ml-en2020/09/18 00:35:003849---0.721277------NMTYes
3cvitINDIC20ml-en2020/10/14 13:00:324146---0.716216------NMTNoBaseline NMT fine-tuned then pre-trained on filtered Cyclic BT corpus further fine-tune.
4HW-TSCINDIC20ml-en2020/09/19 11:47:564042---0.705297------NMTNotransformer deep(pre LN),xx2en,PMI data and filtered PBI data (fasttext domain adaptation)+ sentencepiece + 2 model ensemble
5NICT-5INDIC20ml-en2020/09/18 21:22:264004---0.640107------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size unbalanced.
6ORGANIZERINDIC20ml-en2020/09/02 16:44:303634---0.577294------NMTNoBaseline MLNMT XX to En model using PIB and Filtered PMI data. Transformer big model. Default settings.
7NICT-5INDIC20ml-en2020/09/18 21:01:373991---0.569790------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size balanced.
8ODIANLPINDIC20ml-en2020/09/17 21:41:123821---0.567925------NMTNoTransformer Big with Relative position representations + xx-en model + PMI Data

Notice:

Back to top

AMFM


# Team Task Date/Time DataID AMFM
Method
Other
Resources
System
Description
unuse unuse unuse unuse unuse unuse unuse unuse unuse unuse
1ORGANIZERINDIC20ml-en2020/09/02 16:44:303634---0.000000------NMTNoBaseline MLNMT XX to En model using PIB and Filtered PMI data. Transformer big model. Default settings.
2ODIANLPINDIC20ml-en2020/09/17 21:41:123821---0.000000------NMTNoTransformer Big with Relative position representations + xx-en model + PMI Data
3cvitINDIC20ml-en2020/09/18 00:22:373843---0.000000------NMTYes
4cvitINDIC20ml-en2020/09/18 00:35:003849---0.000000------NMTYes
5NICT-5INDIC20ml-en2020/09/18 21:01:373991---0.000000------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size balanced.
6NICT-5INDIC20ml-en2020/09/18 21:22:264004---0.000000------NMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size unbalanced.
7HW-TSCINDIC20ml-en2020/09/19 11:47:564042---0.000000------NMTNotransformer deep(pre LN),xx2en,PMI data and filtered PBI data (fasttext domain adaptation)+ sentencepiece + 2 model ensemble
8cvitINDIC20ml-en2020/10/14 13:00:324146---0.000000------NMTNoBaseline NMT fine-tuned then pre-trained on filtered Cyclic BT corpus further fine-tune.

Notice:

Back to top

HUMAN (WAT2022)


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

Notice:
Back to top

HUMAN (WAT2021)


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

Notice:
Back to top

HUMAN (WAT2020)


# Team Task Date/Time DataID HUMAN
Method
Other
Resources
System
Description
1ODIANLPINDIC20ml-en2020/09/17 21:41:123821UnderwayNMTNoTransformer Big with Relative position representations + xx-en model + PMI Data
2cvitINDIC20ml-en2020/09/18 00:22:373843UnderwayNMTYes
3NICT-5INDIC20ml-en2020/09/18 21:22:264004UnderwayNMTNoXX to XX transformer model trained on officially provided PMI and PKB data. Corpora were size unbalanced.
4HW-TSCINDIC20ml-en2020/09/19 11:47:564042UnderwayNMTNotransformer deep(pre LN),xx2en,PMI data and filtered PBI data (fasttext domain adaptation)+ sentencepiece + 2 model ensemble

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

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