Hi, this is commonly referred to as "system combination" and is carried out on the output when translating with different models.
See for instance: http://kheafield.com/code/memt/ http://www-lium.univ-lemans.fr/~barrault/MANY -phi On Wed, Jun 18, 2014 at 5:23 AM, Megha Jhunjhunwala <[email protected]> wrote: > Hello, > > I am trying to implement ensemble decoding to combine different models I > have obtained after triangulation using different pivots. I want to know if > there is any predefined function in moses which can be used so that at the > time of decoding I can dynamically combine different models. If there is no > predefined function then please suggest some alternate methods to implement > it. > > Thanks, > > Megha > > _______________________________________________ > 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
