The aim of this research is to establish the role of linguistic information in data-scarce statistical machine translation for sign languages using freely available tools. The main challenge in statistical machine translation is the scarcity of suitable data, and this problem becomes more pronounced in sign languages. The available corpora are small, usually not domain-specific, and their annotation conventions can vary considerably. Elaborating our own corpus is a very time-consuming task and the amount of data that we can obtain is even more reduced. Under these conditions, morpho-syntactic information helps to improve statistical machine translation results, but there are not linguistic processing tools for sign languages. We have managed to improve translations from Catalan to Catalan Sign Language by using factored models in an open source translation system with basic linguistic information such as the lemma or an annotation tier tag. Furthermore, this allows us to deal with sign language morphemes in a more systematic way.
@inproceedings{masso:10004:sign-lang:lrec,
author = {Mass{\'o}, Guillem and Badia, Toni},
title = {Dealing with Sign Language Morphemes in Statistical Machine Translation},
pages = {154--157},
editor = {Dreuw, Philippe and Efthimiou, Eleni and Hanke, Thomas and Johnston, Trevor and Mart{\'i}nez Ruiz, Gregorio and Schembri, Adam},
booktitle = {Proceedings of the {LREC2010} 4th Workshop on the Representation and Processing of Sign Languages: Corpora and Sign Language Technologies},
maintitle = {7th International Conference on Language Resources and Evaluation ({LREC} 2010)},
publisher = {{European Language Resources Association (ELRA)}},
address = {Valletta, Malta},
day = {22--23},
month = may,
year = {2010},
language = {english},
url = {https://www.sign-lang.uni-hamburg.de/lrec/pub/10004.pdf}
}