摘要 |
According to one embodiment, a machine translation apparatus includes a translation unit, an acquisition unit, a first calculation unit, a reverse translation unit, a second calculation unit and a selection unit. The acquisition unit acquires, at least one second forward-translated word different from the first forward-translated word, to obtain candidate words. The reverse translation unit obtains at least one reverse-translated word for each of the candidate words by reverse-translating each candidate word into the first language. The selection unit selects a corrected forward-translated word to be replaced with the first forward-translated word from among the candidate words based on the semantic similarity and fluency. |