Send Moses-support mailing list submissions to
moses-support@mit.edu
To subscribe or unsubscribe via the World Wide Web, visit
http://mailman.mit.edu/mailman/listinfo/moses-support
or, via email, send a message with subject or body 'help' to
moses-support-request@mit.edu
You can reach the person managing the list at
moses-support-owner@mit.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Moses-support digest..."
Today's Topics:
1. Call for Participation: WAT2014 (The 1st Workshop on Asian
Translation) (Toshiaki Nakazawa)
2. Re: suffix array, how to specify grammar file per sentence
(Hieu Hoang)
3. Regarding an issue during Translation (Chitra C)
4. Re: Regarding an issue during Translation (Barry Haddow)
----------------------------------------------------------------------
Message: 1
Date: Tue, 13 May 2014 13:08:41 +0900
From: Toshiaki Nakazawa <nakazawa@pa.jst.jp>
Subject: [Moses-support] Call for Participation: WAT2014 (The 1st
Workshop on Asian Translation)
To: moses-support@mit.edu
Message-ID: <m2r43ys4l2.wl%nakazawa@pa.jst.jp>
Content-Type: text/plain; charset=US-ASCII
Dear all,
I'm Toshiaki Nakazawa from JST (Japan Science and Technology Agency),
Japan. This is an announcement of the new open machine translation
evaluation campaign. Those who are working on machine translation,
please join us.
Best regards,
---------------------------------------------------------------------------
WAT 2014
(The 1st Workshop on Asian Translation)
http://orchid.kuee.kyoto-u.ac.jp/WAT/
October 4, 2014, Tokyo, Japan
WAT is a new open evaluation campaign focusing on Asian languages. We
would like to invite a broad range of participants and conduct various
forms of machine translation experiments and evaluation. Collecting
and sharing our knowledge will allow us to understand the essence of
machine translation and the problems to be solved. We are working
toward the practical use of machine translation among all Asian
countries.
For the 1st WAT, we chose "scientific papers" as the targeted domain,
and selected the languages Japanese, Chinese and English.
TASK
----
The task is to improve the text translation quality for scientific
papers. Participants choose any of the subtasks in which they would
like to participate and translate the test data using their machine
translation systems. The WAT organizers will evaluate the results
submitted using automatic evaluation and human evaluation. We will
also provide a baseline machine translation.
Subtasks:
Japanese --> English
English --> Japanese
Japanese --> Chinese
Chinese --> Japanese
Dataset:
WAT uses ASPEC (Asian Scientific Paper Excerpt Corpus) for the dataset
including training, development, development test and test
data. Participants must get a copy of ASPEC by themselves from
http://orchid.kuee.kyoto-u.ac.jp/ASPEC/
Automatic evaluation:
We will prepare an automatic evaluation server. You will be able to
evaluate the translation results at any time using this server.
Human evaluation:
Human evaluation will be carried out using crowdsourcing. Participants
can submit translation results a maximum of
twice. Sentence-by-sentence pair-wise evaluation compared to the
baseline system will be carried out. The crowdsourcing workers will be
asked to judge which translation is better than the other in view of
adequacy and fluency. All systems will be ranked by the percentage of
translations judged to improve upon the baseline system.
REGISTRATION
------------
The registration procedure for WAT2014 will be announced
later. However, you can start developing and evaluating your MT
engines today using ASPEC. The registration fee will be FREE for all
participants.
IMPORTANT DATES
---------------
Crowdsourcing evaluation due July 31, 2014
System description draft paper due August 31, 2014
Review feedback September 7, 2014
Camera-ready paper due September 14, 2014
WAT 2014 October 4, 2014
PAPER
-----
Participants who submit results for human evaluation should submit
description papers of their translation systems and evaluation
results.
We strongly prefer that papers include a section entitled "Issues for
Context-aware Machine Translation" which discusses the importance and
usefulness of context.
ORGANIZERS
----------
Toshiaki Nakazawa (Japan Science and Technology Agency (JST))
Hideya Mino (National Institute of Information and Communications Technology (NICT))
Isao Goto (Japan Broadcasting Corporation (NHK))
Eiichiro Sumita (National Institute of Information and Communications Technology (NICT))
Sadao Kurohashi (Kyoto University)
CONTACT
-------
wat@nlp.ist.i.kyoto-u.ac.jp
---------------------------------------------------------------------------
--
Toshiaki Nakazawa (Researcher)
Japan Science and Technology Agency (JST)
(@ Graduate School of Informatics, Kyoto University)
Yoshida-honmachi, Sakyo-ku, Kyoto, 606-8501, Japan
tel: +81-75-753-5346, fax: +81-75-753-5962
nakazawa@pa.jst.jp / nakazawa@nlp.ist.i.kyoto-u.ac.jp
------------------------------
Message: 2
Date: Tue, 13 May 2014 10:33:11 +0100
From: Hieu Hoang <Hieu.Hoang@ed.ac.uk>
Subject: Re: [Moses-support] suffix array, how to specify grammar file
per sentence
To: Mohammad Salameh <msalameh83@gmail.com>
Cc: moses-support <moses-support@mit.edu>
Message-ID:
<CAEKMkbhA0XgntfWoB64hCnmj50Ntna0G7mNPv-h3V75yfPbgCQ@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"
not really. You can specify
-start-translation-id 555
to the decoder and the grammar file that it uses will be grammar.555,
grammar.556 ....
The sequence file that contains
<seg grammar=
looks like something that cdec would understand
On 11 May 2014 03:13, Mohammad Salameh <msalameh83@gmail.com> wrote:
> Hi All,
> I am using suffix array for hierarichal models. And I have generated the
> grammar per sentence files using
> ./scripts/training/wrappers/adam-suffix-array/suffix-array-create.sh
>
> During decoding with moses_chart I am providing the directory path in the
> config file for all the grammar sentences, which works fine. But is there
> a way that I can specify which grammar file I want to use for each sentence
> in the test file ?
>
> The script above also generates a sentence file of the format:
> <seg grammar="/local/scratch/work/a4/evaluation/ORD/grammar.0.gz"
> id="0">erdogan states turkey to reject any pressures to urge it to
> recognize cyprus </seg>
>
> but the decoder seems not to be aware of the path in the sentence.
>
> Regards,
> Salameh
>
>
>
>
> _______________________________________________
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>
--
Hieu Hoang
Research Associate
University of Edinburgh
http://www.hoang.co.uk/hieu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20140513/6845e974/attachment-0001.htm
------------------------------
Message: 3
Date: Tue, 13 May 2014 06:40:48 -0400
From: Chitra C <chitrac03@gmail.com>
Subject: [Moses-support] Regarding an issue during Translation
To: moses-support@mit.edu
Message-ID:
<CACDiBTHdKdiJvLr1OBd_F-7vBa35EqEuD76jk-dE07UN0C6Dbg@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"
Hello Moses-Support team,
My name is Chitra, I am working on SMT[ Statistical Machine Translation ]
from English to Kannada Language. I am using Giza,IRSTLM and Moses tools.
I am following Oses manual for installation and training.Installation of
all the tools are successful,but during the training of language model i
have a problem.
when i execute the following command
~/irstlm/bin/compile-lm \
--text yes \
kannada.lm.kn.gz \
kannada.arpa.kn
I am getting a warning called-Too many arguments
In order to avoid this, i gave following command
~/irstlm/bin/compile-lm \
--text \
kannada.lm.kn.gz \
kannada.arpa.kn
arpa file got created,but i am getting error when i try to do
tuning,testing ..etc.
The error was like
Loading the LM will be faster if you build a binary file.
Reading kannada.blm.kn
----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100
****************************************************************************************************
lm/read_arpa.cc:63 in void lm::ReadARPACounts(util::FilePiece&,
std::vector<long long unsigned int, std::allocator<long long unsigned int>
>&) threw FormatLoadException because `line.size() >= 4 &&
StringPiece(line.data(), 4) == "blmt"'.
This looks like an IRSTLM binary file. Did you forget to pass --text yes
to compile-lm? Byte: 40 File: kannada.blm.kn
Please help me out to solve this problem,
Thank you in advance.
Regards,
Chitra C
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20140513/b23b310c/attachment-0001.htm
------------------------------
Message: 4
Date: Tue, 13 May 2014 14:05:54 +0100
From: Barry Haddow <bhaddow@staffmail.ed.ac.uk>
Subject: Re: [Moses-support] Regarding an issue during Translation
To: Chitra C <chitrac03@gmail.com>, moses-support@mit.edu
Message-ID: <537218B2.7020303@staffmail.ed.ac.uk>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Hi Chitra
Can you check that you are using the version of IRSTLM specified in the
"baseline system" (http://www.statmt.org/moses/?n=Moses.Baseline) page?
The "--text yes" argument should work.
Also, the "baseline system" page also explains how to create a kenlm
binary file from an ARPA file. It looks like you are passing an irstlm
binary file to kenlm, and that is why Moses is failing.
cheers - Barry
On 13/05/14 11:40, Chitra C wrote:
> Hello Moses-Support team,
>
> My name is Chitra, I am working on SMT[ Statistical Machine
> Translation ] from English to Kannada Language. I am using Giza,IRSTLM
> and Moses tools.
>
> I am following Oses manual for installation and training.Installation
> of all the tools are successful,but during the training of language
> model i have a problem.
>
> when i execute the following command
>
> ~/irstlm/bin/compile-lm \
> --text yes \
> kannada.lm.kn.gz \
> kannada.arpa.kn <http://kannada.arpa.kn>
>
> I am getting a warning called-Too many arguments
> In order to avoid this, i gave following command
>
> ~/irstlm/bin/compile-lm \
> --text \
> kannada.lm.kn.gz \
> kannada.arpa.kn <http://kannada.arpa.kn>
>
> arpa file got created,but i am getting error when i try to do
> tuning,testing ..etc.
>
> The error was like
>
> Loading the LM will be faster if you build a binary file.
> Reading kannada.blm.kn <http://kannada.blm.kn>
> ----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100
> ****************************************************************************************************
> lm/read_arpa.cc:63 in void lm::ReadARPACounts(util::FilePiece&,
> std::vector<long long unsigned int, std::allocator<long long unsigned
> int> >&) threw FormatLoadException because `line.size() >= 4 &&
> StringPiece(line.data(), 4) == "blmt"'.
> This looks like an IRSTLM binary file. Did you forget to pass --text
> yes to compile-lm? Byte: 40 File: kannada.blm.kn <http://kannada.blm.kn>
>
> Please help me out to solve this problem,
> Thank you in advance.
>
> Regards,
> Chitra C
>
>
> _______________________________________________
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
------------------------------
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support
End of Moses-support Digest, Vol 91, Issue 17
*********************************************
Subscribe to:
Post Comments (Atom)
0 Response to "Moses-support Digest, Vol 91, Issue 17"
Post a Comment