Sampling Phrase Tables for the Moses Statistical Machine Translation System
Ulrich Germann · The Prague Bulletin of Mathematical Linguistics · 2015
Abstract The idea of virtual phrase tables for statistical machine translation (SMT) that construct phrase table entries on demand by sampling a fully indexed bitext was first proposed ten years ago by Callison-Burch et al. (2005). However, until recently (Germann, 2014) no working and practical implementation of this approach was available in the Moses SMT system. We describe and evaluate this implementation in more detail. Sampling phrase tables are much faster to build and are competitive with conventional phrase tables in terms of translation quality and speed.