Principal investigator (ÚFAL): 
Provider: 
Grant id: 
639012
Duration: 
2012-2013

Tools and data for Machine Translation between Related Languages

The project aims at creating a suitable formal and applicational framework for machine translation systems among related languages, especially to find out which MT paradigm, rule-based or statistical, will score best for closely-related languages. To evaluate the framework, an experimental machine translation system from Czech to Russian will be implemented within both a rule-based system Česílko and statistical system Moses. The quality of the translation of both systems in various configurations will be compared using BLEU score as well as a manual evaluation metric.

Publications

Klyueva Natalia: Usage of some non-finite constructions in Czech and Russian. In: 6th Annual International Conference on Languages & Linguistics, Copyright © Atiner, Athens, Greece, ISSN 2241-2891, pp. 5-12, 2013


Bílek Karel, Zeman Daniel: CUni Multilingual Matrix in the WMT 2013 Shared Task. In: Proceedings of the Eight Workshop on Statistical Machine Translation, Copyright © Association for Computational Linguistics, Sofija, Bulgaria, ISBN 978-1-937284-57-2, pp. 85-91, 2013
Bílek Karel, Klyueva Natalia, Kuboň Vladislav: Exploiting Maching Learning for Automatic Semantic Feature Assignment. [Draft] Proceedings of the Twenty-Sixth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2013, Copyright © AAAI Press, Palo Alto, California, ISBN 978-1-57735-605-9, 2013
Klyueva Natalia: Some differences between Czech and Russian: a parallel corpus study. In: Компьютерная лингвистика и интеллектуальные технологии, Vol. Issue 11 (18), Computational Linguistics and Intellectual Technologies: Proceedings of the International Conference "Dialog 2012", Copyright © Изд-во РГГУ, Москва, Russia, ISSN 2221-7932, pp. 268-276, 2013

Klyueva Natalia, Bojar Ondřej, Garabík Radovan, Týnovský Miroslav: Czech-Russian corpus via a simple web interface. Contributed talk, Workshop on parallel corpora, Mainz, Germany, Sep 2012 (oral presentation)
Klyueva Natalia: Comparing Czech and Russian Valency on the Material of Vallex. In: Empirical Methods in Natural Language Processing - Proceedings of the Conference on Natural Language Processing 2012, Copyright © Eigenverlag ÖGAI, Wien, Austria, ISBN 3-85027-005-X, pp. 446-451, 2012