Moses-support Digest, Vol 124, Issue 5

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. Special Session on Multilingual Processing in International
Conference in Dehradun. (Vishal Goyal (????? ????))
2. Re: BLEU score decoding word lattice (Hieu Hoang)
3. Job opportunity: highly skilled MT engineer in WIPO - Geneva
(Pouliquen, Bruno)
4. Re: Sought Help in resolving tuning error using mert-moses.pl
command (Hieu Hoang)


----------------------------------------------------------------------

Message: 1
Date: Fri, 3 Feb 2017 12:57:35 +0530
From: Vishal Goyal(????? ????) <vishal.pup@gmail.com>
Subject: [Moses-support] Special Session on Multilingual Processing in
International Conference in Dehradun.
To: moses-support <moses-support@mit.edu>
Message-ID:
<CADAAq5yitwt9txze0YgBheaN_7q1hc8Y0ARb436AgQML5-_Mag@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Dear All,

Please find the details at following web link -

http://www.icacca.in/special-session-on-multilingual-processing/



--
*Regards,*
Vishal Goyal,
Ph.D., M.Tech., MCA, M.C.S.D.
Associate Professor(Stage IV),
Department of Computer Science,
Punjabi University Patiala-147002.


*Machine Translation Systems:*
[*Online Hindi to Punjabi Machine Translation Tool -*
http://h2p.learnpunjabi.org ]
[*Statistical Approach Based Hindi to Punjabi Machine Translation System *
- http://statmt.org/~vishal/hp/index.cgi
- http://tdil-dc.in/hi2pu/index.cgi
]
*Online Journal: [Research Cell: An International Journal of Engineering
Sciences, http://ijoes.vidyapublications.com
<http://ijoes.vidyapublications.com>]*
*Book: A Simplified Approach to Data Structures, Shroff Publications and
Distributors*
http://www.shroffpublishers.com/detail.aspx?title=6163
*Theory Of Computation Video Lectures on YouTube:* https://www.youtube.
com/channel/UCWkhzvxKuJ5G2SF2opeGngg



--
*Regards,*
Vishal Goyal,
Ph.D., M.Tech., MCA, M.C.S.D.
Associate Professor(Stage IV),
Department of Computer Science,
Punjabi University Patiala-147002.


*Machine Translation Systems:*
[*Online Hindi to Punjabi Machine Translation Tool -*
http://h2p.learnpunjabi.org ]
[*Statistical Approach Based Hindi to Punjabi Machine Translation System *
- http://statmt.org/~vishal/hp/index.cgi
- http://tdil-dc.in/hi2pu/index.cgi
]
*Online Journal: [Research Cell: An International Journal of Engineering
Sciences, http://ijoes.vidyapublications.com
<http://ijoes.vidyapublications.com>]*
*Book: A Simplified Approach to Data Structures, Shroff Publications and
Distributors*
http://www.shroffpublishers.com/detail.aspx?title=6163
*Theory Of Computation Video Lectures on YouTube:*
https://www.youtube.com/channel/UCWkhzvxKuJ5G2SF2opeGngg
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20170203/79aa749a/attachment-0001.html

------------------------------

Message: 2
Date: Fri, 3 Feb 2017 10:12:52 +0000
From: Hieu Hoang <hieuhoang@gmail.com>
Subject: Re: [Moses-support] BLEU score decoding word lattice
To: Angli Liu <anglil@cs.washington.edu>
Cc: moses-support <moses-support@mit.edu>
Message-ID:
<CAEKMkbgnWs6wN2NVYMg5XLqY0NS-sPmrW2hOHtfXXr=7fmENWw@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

yes, the phrase-based decoder supports lattice decoding
http://www.statmt.org/moses/?n=Moses.WordLattices

Hieu Hoang
http://moses-smt.org/

On 1 February 2017 at 23:10, Angli Liu <anglil@cs.washington.edu> wrote:

> Hi all,
>
> Is there a way to do lattice decoding with BLEU in Moses? I.e., given a
> word lattice, find the path that represents the highest BLEU score? If so,
> what function to call and in what format should I feed a lattice in?
>
> Thanks!
> Angli
>
> _______________________________________________
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20170203/6441fc05/attachment-0001.html

------------------------------

Message: 3
Date: Fri, 3 Feb 2017 11:03:46 +0000
From: "Pouliquen, Bruno" <bruno.pouliquen@wipo.int>
Subject: [Moses-support] Job opportunity: highly skilled MT engineer
in WIPO - Geneva
To: "moses-support@mit.edu" <moses-support@mit.edu>
Message-ID: <48496c8bf16c4d119f5a20ae51212b9b@WICM02.WIECSP.UNICC.ORG>
Content-Type: text/plain; charset="us-ascii"

Title: R&D Engineer needed in SMT and NMT
The World Intellectual Property Organization in Geneva (WIPO - Geneva Switzerland) is looking for a highly skilled person in the area of statistical machine translation, with knowledge of neural machine translation techniques.
More information is available online at: https://wipo.taleo.net/careersection/wp_03/jobsearch.ftl?lang=en
Context:
WIPO Translate has been working in the area of automatically translating Patent documents using open source Moses for the last seven years.
In September 2016, thanks to our collaborator Marcin Junczys-Dowmunt and his open-source AmuNMT, we put Neural MT in production (http://www.wipo.int/patentscope/en/wipo-translate/).
We plan to extend Neural MT to other languages very soon, and extend it also to other services of WIPO and other international organizations.
Position:
This is a short term position (one year, with the possibility of a maximum one year extension), with an attractive salary. Located in Geneva.
The candidate must have a strong background in MT (hopefully in Neural Machine Translation - Nematus / AmuNMt) with the ability to maintain and extend existing software and the desire to take the latest research technologies into production.
Note the very short deadline: 16/2/2017
Please apply online at: https://wipo.taleo.net/careersection/wp_03/jobsearch.ftl?lang=en



World Intellectual Property Organization Disclaimer: This electronic message may contain privileged, confidential and copyright protected information. If you have received this e-mail by mistake, please immediately notify the sender and delete this e-mail and all its attachments. Please ensure all e-mail attachments are scanned for viruses prior to opening or using.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20170203/00ce0b24/attachment-0001.html

------------------------------

Message: 4
Date: Fri, 3 Feb 2017 13:28:04 +0000
From: Hieu Hoang <hieuhoang@gmail.com>
Subject: Re: [Moses-support] Sought Help in resolving tuning error
using mert-moses.pl command
To: Arpita Dutta <duttaarpi@gmail.com>, moses-support
<moses-support@mit.edu>
Message-ID:
<CAEKMkbh=m8gS-CRmB_QtnU_uhwbSS=efAcTnzO=aUNJnBXdfwA@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Please ask question on the moses mailing list rather than individual people
directly. Then others can answer and learn from your question. Don't forget
to subscribe to the mailing list 1st before asking, you can subscribe here:
http://mailman.mit.edu/mailman/listinfo/moses-support

To answer your question - yes, you must run tuning after you create the
moses.ini file in the model folder.

The decoder failed to run. This could be because it hasn't compiled
successfully, or you've mispelt the folder where it's supposed to be

Hieu Hoang
http://moses-smt.org/

On 3 February 2017 at 13:11, Arpita Dutta <duttaarpi@gmail.com> wrote:

> Respected Sir,
>
> I would like to introduce myself as a research scholar from
> India,working in machine translation from english to bengali in National
> Institute of Technology Silchar, India (www.nits.ac.in).
>
> Sir, I am following Prof. Koehen's Moses manual for decoding purpose where
> I am facing the error while running the code for tuning given below.The
> error is highlighted.
>
>
> nits@cse:~/Desktop/ad/demo$ /home/nits/Desktop/ad/
> mosesdecoder/scripts/training/mert-moses.pl /home/nits/Desktop/ad/demo/data/tuning.en
> /home/nits/Desktop/ad/demo/data/tuning.ben /home/nits/Desktop/ad/mosesdecoder/bin/moses
> /home/nits/Desktop/ad/demo/model/moses.ini --working-dir
> /home/nits/Desktop/ad/demo/tuning --rootdir /home/nits/Desktop/ad/
> mosesdecoder/scripts
> Using SCRIPTS_ROOTDIR: /home/nits/Desktop/ad/mosesdecoder/scripts
> Assuming --mertdir=/home/nits/Desktop/ad/mosesdecoder/bin
> filtering the phrase tables... Fri Feb 3 18:19:54 IST 2017
> exec: /home/nits/Desktop/ad/mosesdecoder/scripts/training/
> filter-model-given-input.pl ./filtered /home/nits/Desktop/ad/demo/model/moses.ini
> /home/nits/Desktop/ad/demo/data/tuning.en
> Executing: /home/nits/Desktop/ad/mosesdecoder/scripts/training/
> filter-model-given-input.pl ./filtered /home/nits/Desktop/ad/demo/model/moses.ini
> /home/nits/Desktop/ad/demo/data/tuning.en > filterphrases.out 2>
> filterphrases.err
> Asking moses for feature names and values from filtered/moses.ini
> Executing: /home/nits/Desktop/ad/mosesdecoder/bin/moses -config
> filtered/moses.ini -show-weights
> exec: /home/nits/Desktop/ad/mosesdecoder/bin/moses -config
> filtered/moses.ini -show-weights
> Executing: /home/nits/Desktop/ad/mosesdecoder/bin/moses -config
> filtered/moses.ini -show-weights > ./features.list 2> /dev/null
> Exit code: 1
> ERROR: Failed to run '/home/nits/Desktop/ad/mosesdecoder/bin/moses
> -config filtered/moses.ini -show-weights'. at /home/nits/Desktop/ad/
> mosesdecoder/scripts/training/mert-moses.pl line 1748.
> nits@cse:~/Desktop/ad/demo$
>
> This moses.ini conguration file has been created after tuning.
>
> #########################
> ### MOSES CONFIG FILE ###
> #########################
>
> # input factors
> [input-factors]
> 0
>
> # mapping steps
> [mapping]
> 0 T 0
>
> [distortion-limit]
> 6
>
> # feature functions
> [feature]
> UnknownWordPenalty
> WordPenalty
> PhrasePenalty
> PhraseDictionaryMemory name=TranslationModel0 num-features=4
> path=/home/nits/Desktop/ad/demo/tuning/filtered/phrase-table.0-0.1.1.gz
> input-factor=0 output-factor=0
> LexicalReordering name=LexicalReordering0 num-features=6
> type=wbe-msd-bidirectional-fe-allff input-factor=0 output-factor=0
> path=/home/nits/Desktop/ad/demo/tuning/filtered/reordering-table.wbe-msd-bidirectional-fe.0-0.1
>
> Distortion
> IRSTLM name=LM0 factor=0 path=/home/nits/Desktop/ad/demo/lm/train.ilm.gz
> order=3
>
> # dense weights for feature functions
> [weight]
> # The default weights are NOT optimized for translation quality. You MUST
> tune the weights.
> # Documentation for tuning is here: http://www.statmt.org/moses/?
> n=FactoredTraining.Tuning
> UnknownWordPenalty0= 1
> WordPenalty0= -1
> PhrasePenalty0= 0.2
> TranslationModel0= 0.2 0.2 0.2 0.2
> LexicalReordering0= 0.3 0.3 0.3 0.3 0.3 0.3
> Distortion0= 0.3
> LM0= 0.5
>
>
>
> *1. Whether my process of tuning is corrector not.If it is wrong ,How to
> tune the weights properly?*
>
> *2.whether I have to tune the weights after creating moses.ini file in
> model folder.*
> Sir please help me out how to resolve this problem.If my code is wrong
> then kindly suggest me the correct code for this.
>
> With Regards
>
> Mrs. Arpita dutta
> Phd Scholar
> NIT Silchar, India
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20170203/c850e187/attachment.html

------------------------------

_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support


End of Moses-support Digest, Vol 124, Issue 5
*********************************************

Tweet