you can use the argument -t when running the decoder. This will change the
output from
   this is a small house
to
   this is |0-1| a |2-2| small |3-3| house |4-4|
where the numeric ranges are the position of the source phrase

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


On 6 November 2017 at 17:44, Denis Kulagin <[email protected]> wrote:

> I have trained Moses model on a large parallel corpora. How do I find a
> translation of a particular phrase within a sentence?
>
> E.g. I have the following source [en] sentence X:
>
> Free media is a _double_edged_sword_ that can give benefits to any nation
> and can hurt by propaganda.
>
> And a target translation Y. How do I find out what part of the target Y
> corresponds to the "double edged sword" phrase from the source sentence X?
>
> _______________________________________________
> 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