Machine translation study granted DKK 1.6 million

New funds for language research

The Danish Research Council for Humanities (FKK) has granted Research Assistant Jakob Elming from the Department of International Language Studies and Computational Linguistics DKK 1.65 million to a research project on ‘syntax-informed statistical machine translation’.

- Statistical machine translation is different from earlier systems in that it teaches itself to translate based on documents, which have been translated by human translators. This means that the technology approximates artificial intelligence, and that opens up for translation systems handling the small languages like Danish, Jakob Elming says.

Statistical machine translation functions at word level, and the purpose of the project is to improve the system by integrating knowledge about sentence structure, ultimately making it possible for the system to provide better translations.

Read more about 2009 grants of the Danish Research Council for Humanities here