Phrase Model Training for Statistical Machine Translation with Word Lattices of Preprocessing Alternatives

28074

Knowledge Triples