Paper: Fast Consensus Decoding over Translation Forests

ACL ID P09-1064
Title Fast Consensus Decoding over Translation Forests
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2009
Authors

The minimum Bayes risk (MBR) decoding ob- jectiveimprovesBLEUscoresformachinetrans- lation output relative to the standard Viterbi ob- jective of maximizing model score. However, MBRtargetingBLEUisprohibitivelyslowtoop- timize over k-best lists for large k. In this pa- per, we introduce and analyze an alternative to MBR that is equally effective at improving per- formance, yet is asymptotically faster — running 80 times faster than MBR in experiments with 1000-best lists. Furthermore, our fast decoding procedure can select output sentences based on distributions over entire forests of translations, in addition to k-best lists. We evaluate our proce- dure on translation forests from two large-scale, state-of-the-art hierarchical machine translation systems. Our forest-based decoding objec...