Hi, the lexicalized reordering model will be helpful, but the verb movements in your case are often over a very long distance, so it alone will not suffice. A successful strategy is our pre-reordering approach (look up the 2005 ACL Collins / Koehn / Kucerova paper), but it does require a parser in the source language.
-phi On Tue, Feb 22, 2011 at 8:17 AM, Abbasov Samir <[email protected]> wrote: > I am working on russian-turkish statistical machine translation, and > as we know that russian is SVO and turkish is SOV language, does > reordering-table can help me? > > On 22 February 2011 12:11, Abbasov Samir <[email protected]> wrote: >> I am working on russian-turkish statistical machine translation, and >> as we know that russian is SVO and turkish is SOV language, does >> reordering-table help me? >> >> On 22 February 2011 12:07, Abbasov Samir <[email protected]> wrote: >>> Hi everyone, where can i widely read about reordering-table? >>> >>> On 21 February 2011 14:43, Abbasov Samir <[email protected]> wrote: >>>> Hi everyone, in phrase-table has five scores, and then end of each >>>> line has two numbers, what do mean these two numbers? Please, explain >>>> me >>>> >>> >> > _______________________________________________ > 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
