Paper: Discriminative Instance Weighting for Domain Adaptation in Statistical Machine Translation

ACL ID D10-1044
Title Discriminative Instance Weighting for Domain Adaptation in Statistical Machine Translation
Venue Conference on Empirical Methods in Natural Language Processing
Session Main Conference
Year 2010
Authors

We describea new approachto SMT adapta- tion that weights out-of-domainphrase pairs accordingto their relevance to the target do- main, determined by both how similar to it they appearto be, and whetherthey belongto general languageor not. This extends previ- ous work on discriminative weightingby us- ing a finer granularity, focusingon the prop- erties of instances rather than corpus com- ponents, and using a simpler training proce- dure. We incorporateinstanceweightinginto a mixture-modelframework, and find that it yields consistent improvements over a wide rangeof baselines.