Moses-support Digest, Vol 105, Issue 11

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. Re: tokenizer.perl weirdness with some patterns (Pidong Wang)
2. first Call for Participation of DL4MT Winter School (Qun Liu)


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

Message: 1
Date: Thu, 2 Jul 2015 13:01:23 -0700
From: Pidong Wang <wpd1hl@gmail.com>
Subject: Re: [Moses-support] tokenizer.perl weirdness with some
patterns
To: Ozan ?a?layan <ozancag@gmail.com>
Cc: "moses-support@mit.edu" <moses-support@mit.edu>
Message-ID:
<CAKYvXDF6LmKZj0GOEZXHp4M+qJ=Ud_BdKfPjM+2RivCTRL409g@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Yes, this is expected. I do not know the exact reason, but I guess we
assume well-written input which has proper casing (e.g., "I don't
understand your reactions. *S*orry.").

Best wishes!
Pidong

On 2 July 2015 at 12:48, Ozan ?a?layan <ozancag@gmail.com> wrote:

> Hello,
>
> $ echo "tu ne peux pas me voir. blabla" | tokenizer.perl -l fr
> tu ne peux pas me voir. blabla
>
> $ echo -n "I don't understand your reactions. sorry." | tokenizer.perl -l
> en
> I don &apos;t understand your reactions. sorry .
>
> So the problem is that if a dot is followed by a space and then a
> lowercase letter, it is not tokenized. This is happening in at least
> the french tasks of IWSLT. Is this expected? The responsible line for
> this problem is tokenizer.perl:330. What should I lose if I comment
> out the responsible part for this in large scale processing?
>
> Thanks.
>
> PS: I also filed an issue for this:
> https://github.com/moses-smt/mosesdecoder/issues/118
>
>
>
> --
> Ozan ?a?layan
> Research Assistant
> Galatasaray University - Computer Engineering Dept.
> http://www.ozancaglayan.com
>
> _______________________________________________
> 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/20150702/3ff0efc4/attachment-0001.htm

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

Message: 2
Date: Sun, 05 Jul 2015 19:01:47 +0800
From: Qun Liu <liuquncn@gmail.com>
Subject: [Moses-support] first Call for Participation of DL4MT Winter
School
To: corpora@uib.no, mt-list@eamt.org, moses-support@mit.edu,
IRList@lists.shef.ac.uk
Message-ID: <55990E9B.5020402@gmail.com>
Content-Type: text/plain; charset="utf-8"

Apologies for multiple postings. Please distribute to colleagues

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

*

First Call For Participation:

DL4MT Winter School

Deep Learning for Machine Translation

*

18-24 October 2015, Dublin City University, Dublin, Ireland
*

http://dl4mt.computing.dcu.ie/


We would like to announce the Deep Learning For Machine Translation
(DL4MT) Winter School at Dublin City University, Dublin, Ireland.


Over the last several years, Deep learning (DL) has been the driving
force behind huge improvements in speech and image processing. This has
led to high expectations for DL in NLP and MT. In recent top
conferences, a significant portion of papers in the MT domain are
related to DL. Some recent publications have shown the effectiveness of
DL in various aspects of statistical MT. However, because of the
complexity of the implementations and lack of enough details in some of
these publications, it is difficult to repeat the work reported in these
papers. Furthermore, we believe that many MT researchers currently lack
the necessary expertise to incorporate DL into their research, despite
having a high-level understanding of the uses of DL in the MT domain.
Most major NLP conferences have included a deep-learning tutorial for
the last 2-3 years. However, existing tutorials typically do not go into
sufficient depth for participants to actually apply DL algorithms in
their MT research. Because this field is evolving very quickly, we
believe that a multi-day training event will help to prepare MT
researchers to delve into DL, and give them the expertise to apply DL to
their own work.


This one-week programme is sponsored by the European Association for
Machine Translation (EAMT), and organised by The ADAPT Centre for
Digital Content Technology..


We are pleased to announce three excellent mentors:

*

Prof. Kevin Duh, NAIST, Japan (will be affiliated with JHU, US
during the winter school)

*

Prof. Hermann Ney, RWTH, Germany

*

Prof. Kyunghyun Cho, New York University, US

who will present talks on various aspects of DL, with a focus on
applications to MT.


The core themes of the DL4MT workshop will cover:

*

The Fundamentals of DL4MT

*

Neural Language Models and Translation Models for SMT

*

Neural MT (Sequence to Sequence MT/Encoding-Decoding Models)


The structure of the DL4MT Winter School will be as follows: morning
lectures will present in-depth interactive tutorials on topics in DL4MT,
while afternoon sessions will be focused on implementation,

and will take place in a collaborative environment, with support from
expert mentors.


We will solicit applications from MT researchers who have already used
DL techniques in their work, and also from researchers who are
interested in using DL, but would like to enhance their understanding of
DL. This summer workshop focuses on applications of DL to MT, and
attendees should leave with a deep understanding of the state-of-the-art
in DL4MT, and the practical knowledge to implement the core algorithms
in this area.


We will announce the application and registration process in the coming
weeks.


Hope to see you in Dublin!



Organization Committee:


Qun Liu

Tsuyoshi Okita

Chris Hokamp

John Judge

Joachim Wagner


Sponsors:


EAMT <http://www.eamt.org/>

ADAPT <http://www.adaptcentre.ie>

EXPERT <http://expert-itn.eu/>

ICHEC <https://www.ichec.ie/>


*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20150705/a20c19c3/attachment.htm

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

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


End of Moses-support Digest, Vol 105, Issue 11
**********************************************

Tweet