WAT 2018

The 5th Workshop on Asian Translation

Baseline Systems

Tools for NMT

For all subtasks

• OpenNMT version 0.9.7
• BPE

For English tokenization

• Moses version 4.0 or Moses version 2.1.1

For Japanese segmentation

• Juman version 7.0

For Chinese segmentation

• Stanford Word Segmenter version 2014-01-04
  (Using Chinese Penn Treebank (CTB) model)

For Korean segmentation

• mecab-ko
  (https://bitbucket.org/eunjeon/mecab-ko/downloads/mecab-0.996-ko-0.9.1.tar.gz)
  (https://bitbucket.org/eunjeon/mecab-ko-dic/downloads/mecab-ko-dic-1.6.1-20140814.tar.gz)
  Install procedure
  

For Hindi segmentation

• Indic NLP Library: For tokenization, normalization and script conversion

For character conversion

• scripts

For extracting texts from recipe JSON files

• scripts

Java

• Version 7 Update 60 (JDK, JRE)
  (For Stanford Word Segmenter)

NICT (National Institute of Information and Communications Technology)
Kyoto University
Last Modified: 2018-07-30