SMT Experiments for Commercial Translation of Subtitles
Abstract
This contribution reports experiments on developing statistical machine translation (SMT) systems of practical use for the professional translation of subtitles. We present results of several methods that were tested for this task, describing both positive and negative outcomes. We believe these results to be of interest for companies considering the integration of SMT in multilingual commercial systems, and researchers interested in the use of current methods for large-scale SMT systems development in a specific domain.
BIB_text
author = {Thierry Etchegoyhen, Mark Fishel, Jie Jiang, Mirajm Sepesy Mauvec},
title = {SMT Experiments for Commercial Translation of Subtitles},
keywds = {
Statistical Machine Translation, Subtitles, Factored Models, Mixture Models
}
abstract = {
This contribution reports experiments on developing statistical machine translation (SMT) systems of practical use for the professional translation of subtitles. We present results of several methods that were tested for this task, describing both positive and negative outcomes. We believe these results to be of interest for companies considering the integration of SMT in multilingual commercial systems, and researchers interested in the use of current methods for large-scale SMT systems development in a specific domain.
}
isbn = {978-3-9524207-0-6},
date = {2013-09-05},
year = {2013},
}
