rsennrich / wmt2014-scripts

scripts used for SMT system submitted to WMT 2014

Scripts for Edinburgh English-German syntax system for WMT 2014 and WMT 2015

This repository contains scripts and an example config used for the Edinburgh syntax submission (UEDIN-SYNTAX) for the English-German shared translation task at the 2014 and 2015 Workshops on Statistical Machine Translation (http://www.statmt.org/wmt14/ http://www.statmt.org/wmt15/).

The scripts will facilitate the reproduction of our results, and may be useful for people who want to use ParZu (or a different parser with the dependency format by Kilian Foth) for SMT, or string-to-tree systems in general. The hybrid compound splitter can also be used for phrase-based systems, and with German as source language.

CONTENTS

• hybrid_compound_splitter.py

  compound splitter for German (hybrid of finite-state and corpus-based methods as described in Fritzinger & Fraser (2010)), with a novel syntactic representation of split compounds for simple compound merging after string-to-tree translation. The syntactic representation of split compounds is treebank independent and described in Sennrich, Williams and Huck (2014).

  The system to WMT 2014 used the following commands for training/applying the compound splitter:

  hybrid_compound_splitter.py -train -syntax -corpus INPUT_FILE -model MODEL_FILE hybrid_compound_splitter.py -write-filler -no-truecase -q -syntax -smor zmorge-{version}-smor_newlemma.a -model MODEL_FILE < INPUT_FILE > OUTPUT_FILE

  In a string-to-tree system with a syntactic representation of compounds, just apply the following regex substitution to the output for compound merging:

  s/ \@(.*?)\@ /\1/g;

• enrich_labelset.py

  modification of ParZu dependency label set for SMT, splitting up overgeneral labels into distinct subtypes. This script can be applied to ParZu output in CONLL format (before conversion into moses format with the script included in mosesdecoder under scripts/training/wrappers/conll2mosesxml.py).

  Use command line option --wmt14 to activate the modifications used for the submission. Assuming you have the (German-side) tokenized corpus as INPUT_FILE, the Moses parsed files are generated as follows:

  /path/to/mosesdecoder/scripts/tokenizer/deescape-special-chars.perl < INPUT_FILE | \
   /path/to/ParZu/parzu -i tokenized_lines --projective | \
   enrich_labelset.py --wmt14 | \
   /path/to/mosesdecoder/scripts/training/wrappers/conll2mosesxml.py
  

• emnlp2015/*

  scripts used for tree binarization, verb particle restructuring, and (a modified) compound splitting. The techniques are described in Sennrich and Haddow (2015).

• example/toy_example*.config

  toy configs for the moses experimental management system (EMS) that document good settings for training string-to-tree system, and automates the integration of ParZu, compound splitting, tuning on a syntactic metric, a relational dependency language model, and other models into the training process. The different toy examples also document our submissions to the WMT 2014/5 shared translation tasks.

  To facilitate reproduction of our results, parses of the German WMT data sets have been released: http://statmt.org/rsennrich/parsed_wmt/

LICENSE

The scripts are available under the LGPL v2.

PUBLICATIONS

The Edinburgh syntax submission to WMT 2014 is described in:

Philip Williams, Rico Sennrich, Maria Nadejde, Matthias Huck, Eva Hasler and Philipp Koehn (2014): Edinburgh's Syntax-Based Systems at WMT 2014. In: Proceedings of the Ninth Workshop on Statistical Machine Translation.

More details are provided in:

Rico Sennrich, Philip Williams, Matthias Huck (2015): A tree does not make a well-formed sentence: Improving syntactic string-to-tree statistical machine translation with more linguistic knowledge. In: Computer Speech & Language, 32(1):27-45. Hybrid Machine Translation: integration of linguistics and statistics.

The Edinburgh syntax submission to WMT 2015 is described in:

Philip Williams, Rico Sennrich, Maria Nadejde, Matthias Huck and Philipp Koehn (2015): Edinburgh's Syntax-Based Systems at WMT 2015. In: Proceedings of the Tenth Workshop on Statistical Machine Translation. Lisbon, Portugal, pp. 199-209.

More details are provided in:

Rico Sennrich (2015): Modelling and Optimizing on Syntactic N-Grams for Statistical Machine Translation. In: Transactions of the Association for Computational Linguistics 3, 169--182.

Rico Sennrich and Barry Haddow (2015): A Joint Dependency Model of Morphological and Syntactic Structure for Statistical Machine Translation. In: Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing. Lisbon, Portugal, pp. 2081-2087.