Paper: Learning Translation Boundaries for Phrase-Based Decoding

ACL ID N10-1016
Title Learning Translation Boundaries for Phrase-Based Decoding
Venue Human Language Technologies
Session Main Conference
Year 2010

Constrained decoding is of great importance notonlyforspeedbutalsofortranslationqual- ity. Previouseffortsexploresoftsyntacticcon- straints which are based on constituent bound- aries deduced from parse trees of the source language. We present a new framework to es- tablish soft constraints based on a more nat- ural alternative: translation boundary rather than constituent boundary. We propose sim- ple classifiers to learn translation boundaries for any source sentences. The classifiers are trained directly on word-aligned corpus with- out using any additional resources. We report the accuracy of our translation boundary clas- sifiers. We show that using constraints based on translation boundaries predicted by our classifiers achieves significant improvements over the baseline on large-sca...