Phrase-based Neural Machine Translation
Current neural machine translation (NMT) often fails in the one-to-many translation of multi-word phrases and collocations. To tackle this problem, phrase-based NMT systems have been proposed; these typically combine word-based NMT with phrase-based statistical MT systems or external phrase dictionaries. These solutions introduce a significant overhead of additional resources and computational costs. In this project, we are working on a phrase-based NMT model built upon continuous-output NMT, in which the decoder generates embeddings of words or phrases.
People
Chan Young Park
PhD'24 @ CMU -> Postdoc @ UW, MSR -> Assistant Professor @ Purdue University