Лексикализованный подход

This paper proposes a novel lexicalized ap- proach for rule selection for syntax-based statistical machine translation (SMT). We build maximum entropy (MaxEnt) mod- els which combine rich context informa- tion …...