Prolog Statistical Machine Translation is a fairly unsophisticated
statistical machine translation program. It consists of a language model
learner (which takes example sentences in the target language and learns
a language model based on trigrams), a dictionary learner (which learns
word-for-word translations), and a search program (which uses the data
from the first two parts to translate a source sentence into the target
language).