Translation as Linear Transduction
Models & Algorithms for Efficient Learning in Statistical Machine Translation

By Markus Saers
February 2011
Uppsala University Press
Distributed by Coronet Books
ISBN: 9789155479763
133 pages, Illustrated
$52.50 Paper original

Automatic translation has seen tremendous progress in recent years, mainly thanks to statistical methods applied to large parallel corpora. Transductions represent a principled approach to modeling translation, but existing transduction classes are either not expressive enough to capture structural regularities between natural languages or too complex to support efficient statisitcal induction on a large scale. A common approach is to severely prune search over a relatively unrestricted space of transductions grammars. These restrictions are often applied at different stages ina pipeline, with the obvious drwaback of committing to irrevocable decisions that should not have been made. In this thesis we will instead restrict the space of transduction grammars to a space that is less expressive, but can be efficiently searched.


Return to Coronet Books main page