The IBM Models allow a target word to be generated by any set of source word indexes (within the distortion limit), which are not required to be contiguous (i.e. a phrase). I don't believe that the current Moses codebase does this, so you would either need to implement it, or live without this capability and only allow target words to be generated by a contiguous sequence of source words. If you're ok with the latter option, you could probably find a way to sneak the IBM Model parameterizations into a phrase table. I leave this as an exercise for the reader.
The ISI Rewrite decoder is designed to work with IBM Model 4, but alas, it appears to be no longer available. You might email the authors to find out. http://www.isi.edu/licensed-sw/rewrite-decoder/ Adam On 16 Jul 2008, at 12:47, Sanne Korzec wrote: > Hi mailing, > > Is there a way to optimize the moses decoder for decoding IBM word > based models? Or is it better to use a different decoder for these > non phrase based models? If so, any suggestion would be helpful. > > Sanne > _______________________________________________ > Moses-support mailing list > [email protected] > http://mailman.mit.edu/mailman/listinfo/moses-support -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
