hello friends,
I use phrase based machine translation but the target structure
is completely different from source like the following

I live in Gonder for  three days   (English source)

Gonder wist lesost kenoch nurialehu    (Amharic target) where the
reordering algorithm is so slow ,which affect the BLUE score of the
model,since the distance of two verbs(source and target is so far), how can
i improve that, since I am run out of time for research end date.


Thanks in Advance,

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

Reply via email to