The recurring translation task of the WMT workshops focuses on news text and (mainly) European language pairs. For this year the language pairs are:
The goals of the shared translation task are:
|Release of training data for shared tasks (by)||31 January, 2018|
|Test data released||May 14, 2018|
|Translation submission deadline||May 21, 2018|
|Start of manual evaluation||June 11, 2018|
|End of manual evaluation||July 2, 2018|
We provide training data for all eight language pairs, and a common framework. The task is to improve current methods. We encourage a broad participation -- if you feel that your method is interesting but not state-of-the-art, then please participate in order to disseminate it and measure progress. Participants will use their systems to translate a test set of unseen sentences in the source language. The translation quality is measured by a manual evaluation and various automatic evaluation metrics. Participants agree to contribute to the manual evaluation about eight hours of work, per system submission.
You may participate in any or all of the eight language pairs. For all language pairs we will test translation in both directions. To have a common framework that allows for comparable results, and also to lower the barrier to entry, we provide a common training set, and a pre-processed version. You are not limited to this training set, and you are not limited to the training set provided for your target language pair.
If you use additional training data (not provided by the WMT18 organisers) or existing translation systems, you must flag that your system uses additional data. We will distinguish system submissions that used the provided training data (constrained) from submissions that used significant additional data resources. Note that basic linguistic tools such as taggers, parsers, or morphological analyzers are allowed in the constrained condition.
Your submission report should highlight in which ways your own methods and data differ from the standard task. You should make it clear which tools you used, and which training sets you used.
The following two aspects of the task are new for 2018.
Another sub-track is learning translation models from monolingual data only. For some existing approaches have a look at this or this paper; unsupervised cross-lingual embedding mapping and lexicon induction can also be a good start (it even has ready code: one, two), but don't let that limit your imagination.
In 2018 the unsupervised task is limited to the constrained task, so only the provided monolingual data is allowed; monolingual Europarl and News Commentary are also to be excluded, since they are largely parallel. The parallel data may neither be used directly to train the systems, nor indirectly to extract or bootstrap lexicons or other parameters. Again, you are free to test this on any language pairs (or to go wild and do unsupervised multilingual translation), with the special highlights being:
At no additional burden on the News Translation Task participants (aside from having to translate much larger input data), we will collectively provide a deeper analysis of various qualities of the translations.
See WMT18 Test Suites Google Document for more details.
Authors of additional test suites will be invited to report on their evaluation method and its results in a separate paper
The data released for the WMT18 new translation task can be freely used for research purposes, we just ask that you cite the WMT18 shared task overview paper, and respect any additional citation requirements on the individual data sets. For other uses of the data, you should consult with original owners of the data sets.
The provided data is mainly taken from public data sources such as the Europarl corpus, and the UN corpus. Additional training data is taken from the News Commentary corpus, which we re-extract every year from the task.
New for 2018: The first release of the ParaCrawl corpus. This is a new crawled corpus for English to Czech, Estonian, Finnish, German and Russian. As this is the first release, it is potentially noisy, but we have observed bleu score increases on older WMT test sets (over a shallow NMT baseline) when using the Czech (+0.6), Finnish (+2.5), Latvian (+0.9) and Romanian (+3.2) versions of ParaCrawl. For German, bleu score dropped by 1.0 (this was with WMT data over-sampled 7 times). Your Mileage May Vary. You may also want to have a look at the corpus filtering task.
We have added suitable additional training data to some of the language pairs.You may also use the following monolingual corpora released by the LDC:
Note that the released data is not tokenized and includes sentences of any length (including empty sentences). All data is in Unicode (UTF-8) format. The following Moses tools allow the processing of the training data into tokenized format:
To evaluate your system during development, we suggest using the 2017 test set. The data is provided in raw text format and in an SGML format that suits the NIST scoring tool. We also release other dev and test sets from previous years.
The 2018 test sets will be created from a sample of online newspapers from August 2017. The English-X sets are created using equal sized samples of English, and language X, with each sample professionally translated into the other language.
We have released development data for the tasks that are new this year, i.e. Estonian-English and Kazakh-English. It is created in the same way as the test set and included in the development tarball.
The news-test2011 set has three additional Czech translations that you may want to use. You can download them from Charles University.
|Europarl v7||628MB||✓||✓||same as previous year, corpus home page|
|Europarl v8||235MB||✓||✓||et-en is new for this year, corpus home page|
|2.8GB||✓||✓||✓||✓||✓||New for 2018 Please use the filtered version.||876MB||✓||✓||✓||Same as last year|
|News Commentary v13||111M||✓||✓||✓||✓||updated|
|CzEng 1.6||3.1GB||✓||Register and download CzEng 1.6. Better results can be obtained by using a subset of sentences, released under a new version name CzEng 1.7.|
|9.1MB||✓||✓||Provided by CMU..|
|44 MB||✓||Distributed by OPUS|
|3.6 GB||✓||✓||Register and download|
|156 MB||✓||✓||✓||Prepared by Tilde|
|Common Crawl||10.5GB||102GB||103 GB||4.0GB||5.3GB||TBC||42GB||18GB||33GB||Deduplicated with development and evaluation sentences removed. English was updated 31 January 2016 to remove bad UTF-8. Downloads can be verified with SHA512 checksums. More English is available for unconstrained participants.|
|BigEst Estonian corpus||1.4GB||Prepared by the University of Tartu, consists mostly of newspaper and journal articles||News Crawl: articles from 2007||3.7MB||92MB||198MB||302MB||
Extracted article text from various online news publications.
The data sets from 2007-2016 (except Estonian), are the same as last year's.
The data sets from 2017 are fully de-duplicated (Estonian).
|News Crawl: articles from 2008||191MB||313MB||672MB||2.3MB||1.5GB|
|News Crawl: articles from 2009||194MB||296MB||757MB||5.1MB||1.6GB|
|News Crawl: articles from 2010||107MB||135MB||345MB||2.5MB||727MB|
|News Crawl: articles from 2011||389MB||746MB||784MB||564MB||3.1GB|
|News Crawl: articles from 2012||337MB||946MB||751MB||568MB||3.1GB|
|News Crawl: articles from 2013||395MB||1.6GB||1.1GB||730MB||4.3GB|
|News Crawl: articles from 2014||380MB||2.1GB||1.4GB||34MB||52MB||801MB||5.3GB (excludes et)|
|News Crawl: articles from 2015||360MB||2.2GB||1.3GB||43MB||203MB||608MB||4.8G (excludes et)|
|News Crawl: articles from 2016||252MB||1.6GB||1GB||34MB||163MB||418MB||77MB||3.7G (excludes et)|
|News Crawl: articles from 2017||323M||1.8GB||1.3GB||36MB||143MB||504MB||135MB||4.2G|
|News Discussions. Version 1 from 2014/15||1.7GB||Extracted from comment sections of online newspapers. Version 3 is new for this year.|
|News Discussions. Version 2 from 2015/16||6.3GB|
|News Discussions. Version 3 from 2017||3.4GB|
The Common Crawl monolingual data is hosted by Amazon Web Services as a public data set. The underlying S3 URL is
To submit your results, please first convert into into SGML format as required by the NIST BLEU scorer, and then upload it to the website matrix.statmt.org.
For Chinese output, you should submit unsegmented text, since our primary measure is human evaluation. For automatic scoring (in the matrix) we use BLEU4 computed on characters, scoring with v1.3 of the NIST scorer only. A script to convert a Chinese SGM file to characters can be found here.
Each submitted file has to be in a format that is used by standard scoring scripts such as NIST BLEU or TER.
This format is similar to the one used in the source test set files that were released, except for:
<tstset trglang="en" setid="newstest2017" srclang="any">, with trglang set to either
ru. Important: srclang is always
The script wrap-xml.perl makes the conversion of a output file in one-segment-per-line format into the required SGML file very easy:
wrap-xml.perl LANGUAGE SRC_SGML_FILE SYSTEM_NAME < IN > OUT
wrap-xml.perl en newstest2018-src.de.sgm Google < decoder-output > decoder-output.sgm
Upload happens in three easy steps:
If you are submitting contrastive runs, please submit your primary system first and mark it clearly as the primary submission.
Evaluation will be done both automatically as well as by human judgement.