Improved tree-to-string transducer for machine translation

Abstract

We propose three enhancements to the tree-to-string (TTS) transducer for machine translation: first-level expansion-based normalization for TTS templates, a syntactic alignment framework integrating the insertion of unaligned target words, and subtree-based ngram model addressing the tree decomposition probability. Empirical results show that these methods improve the performance of a TTS transducer based on the standard BLEU-4 metric. We also experiment with semantic labels in a TTS transducer, and achieve improvement over our baseline system.