Hi Bharani, there are a number of alternative ways to do it, but I think that Moses can not perform reordering on the preprocessing step. You can use parse trees, alignment information, or POS tags to learn a non-deterministic model able to monotonize your corpus. You may want to have a look at: (Popovic and Ney, 2006), (Costa-Jussa and Fonollosa, 2006), (Khalilov et al, 2009)
_______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
