Minimum error-rate training in statistical machine translation using SVMs
J. González-Rubio, D. Ortiz-Martinez and Francisco Casacuberta
In: 4th Iberian Conference on Pattern Recognition and Image Analysis, June 2009, Póvoa de Varzim, Portugal.

Abstract

Different works on training of log-linear interpolation models for statistical machine translation reported performance improvements by optimizing parameters with respect to translation quality, rather than to likelihood oriented criteria. This work presents an alternative mini- mum error-rate training procedure based on structural support vector machines (SSVMs) for log-linear interpolation models which is not lim- ited to the model scaling factors and needs only few iterations to con- verge. Experimental results are reported on the Spanish­English Eu- roparl corpus.

[Edit]