There is in fact an upcoming paper in NAACL that addresses this exact question:

Stream-based Translation Models for Statistical Machine Translation.
Abby Levenberg, Chris Callison-Burch, and Miles Osborne, 2010. In
Proceedings NAACL-2010.

On Fri, Mar 26, 2010 at 12:14 PM, Germán Sanchis Trilles
<[email protected]> wrote:
> Hi Nati,
>
> I'm afraid that what you are asking is an open research (and currently
> quite active) problem, namely, how to build incremental translation
> models. As far as I know, there are a couple of papers suggesting
> directions of how to confront this problem, although such papers are
> more "work in progress" than really established state of the art.
>
> Hence, the only thing I would suggest you to do is to train a new
> translation model with the whole lot, since with such corpora sizes it
> will only take you about a day.
>
> Good luck,
>
> Germán
>
>
>
> Quoting nati g <[email protected]>:
>
>> Hi,
>> could some one pelase help me out.
>>
>> On Wed, Mar 17, 2010 at 1:48 PM, nati g <[email protected]> wrote:
>>
>>> Hi,
>>>
>>>  i had trined a translation model with coprus consiting around half a
>>> million lines.i am able to translate new strings suceesfully.
>>> now i got corpus from another source having  .2 million lines. Now do i
>>> need to traine translation again by adding the coprus
>>> or do we have any other ways to incoorperate translation model from new
>>> corpus to the old once.
>>>
>>> Thanks in advance.
>>>
>>
>
>
>
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
>
>
> _______________________________________________
> Moses-support mailing list
> [email protected]
> http://mailman.mit.edu/mailman/listinfo/moses-support
>

_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to