Paper: Alignment By Agreement

ACL ID N06-1014
Title Alignment By Agreement
Venue Human Language Technologies
Session Main Conference
Year 2006

We present an unsupervised approach to symmetric word alignment in which two simple asymmetric models are trained jointly to maximize a combination of data likelihood and agreement between the models. Compared to the stan- dard practice of intersecting predictions of independently-trained models, joint train- ing provides a 32% reduction in AER. Moreover, a simple and efficient pair of HMM aligners provides a 29% reduction in AER over symmetrized IBM model 4 predictions.