Agreement Constraints for Statistical Machine Translation into German

Languages with rich inflectional morphology pose a difficult challenge for statistical machine translation. To address the problem of morphologically inconsistent output, we add unification-based constraints to the target-side of a string-to-tree model. By integrating constraint evaluation into the decoding process, implausible hypotheses can be penalised or filtered out during search. We use a simple heuristic process to extract agreement constraints for German and test our approach on an English-German system trained on WMT data, achieving a small improvement in translation accuracy as measured by BLEU.