Hypothesis Refinement Using Agreement Constraints in Machine Translation

Phrase-based machine translation like other data driven approaches, are often plagued by irregularities in the translations of words in morphologically rich languages. The phrase-pairs and the language models are unable to capture the long range dependencies which decide the inflection. This paper makes the first attempt at learning constraints between the language-pair where, the target language lacks rich linguistic resources, by automatically learning classifiers that prevent implausible phrases from being part of decoding and at the same time adds consistent phrases. The paper also shows that this approach improves translation quality on the English-Hindi language pair.