You can use my version of giza++ (http://code.google.com/p/inc-giza-pp/) or
mgiza
to align a batch of new data only using the statistics from a prior run.
More details
are here: http://www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc27

Abby

On Thu, Dec 8, 2011 at 9:29 AM, Hieu Hoang <[email protected]> wrote:

> Hi all
>
> Is it possible to take an existing giza/mgiza model and run it on new
> parallel data to get the alignment?
>
> Or is it possible to use the model created by giza/mgiza and use Abby
> Levenberg's incremental giza to create a word alignment on new data?
>
> Thanks
> Hieu
>
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to