Hi, On Mon, May 21, 2012 at 8:09 PM, Yared Mekuria <[email protected]> wrote: > -is there other mechanisms in phrase based SMT, that can perform better than > distortion > probability in reordering of target language?
There is the lexicalized reordering model that usually gives better results. A fall-back method is to use a source side parser and a few hand-built rules to reorder the source that it matches closer the target. > -can Moses work with tagged and morphological analyzed data? and is it more > advantageous? Yes, this may be useful on the source (for backoff to translation from lemmas) and at the target (for use in morphological tag sequence models). Please take a closer look at: http://www.statmt.org/moses/?n=Moses.FactoredTutorial -phi _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
