Pre-reordering for Statistical Machine Translation of Non-fictional Subtitles

This paper describes the challenges of building a Statistical Machine Translation (SMT) system for non-fictional subtitles. Since our experiments focus on a “difficult“ translation direction (i.e. FrenchGerman), we investigate several methods to improve the translation performance. We also compare our in-house SMT systems (including domain adaptation and pre-reordering techniques) to other SMT services and show that prereordering alone significantly improves the baseline systems.