EU-BRIDGE MT: Combined Machine Translation

Proceedings of the Ninth Workshop on Statistical Machine Translation, pages 105–113, Baltimore, Maryland USA, June 26–27, 2014. c⃝2014 Association for Computational Linguistics EU-BRIDGE MT: Combined Machine Translation ∗Markus Freitag, ∗Stephan Peitz, ∗Joern Wuebker, ∗Hermann Ney, ‡Matthias Huck, ‡Rico Sennrich, ‡Nadir Durrani, ‡Maria Nadejde, ‡Philip Williams, ‡Philipp Koehn, †Teresa Herrmann, †Eunah Cho, †Alex Waibel ∗RWTH Aachen University, Aachen, Germany ‡University of Edinburgh, Edinburgh, Scotland †Karlsruhe Institute of Technology, Karlsruhe, Germany ∗{freitag,peitz,wuebker,ney}@cs.rwth-aachen.de ‡{mhuck,ndurrani,pkoehn}@inf.ed.ac.uk ‡v1rsennr@staffmail.ed.ac.uk ‡maria.nadejde@gmail.com,p.j.williams-2@sms.ed.ac.uk †{teresa.herrmann,eunah.cho,alex.waibel}@kit.edu Abstract This paper describes one of the col- laborative efforts within EU-BRIDGE to further advance the state of the art in machine translation between two Euro- pean language pairs, German→English and English→German. Three research institutes involved in the EU-BRIDGE project combined their individual machine translation systems and participated with a joint setup in the shared translation task of the evaluation campaign at the ACL 2014 Eighth Workshop on Statistical Machine Translation (WMT 2014). We combined up to nine different machine translation engines via system combina- tion. RWTH Aachen University, the Uni- versity of Edinburgh, and Karlsruhe In- stitute of Technology developed several individual systems which serve as sys- tem combination input. We devoted spe- cial attention to building syntax-based sys- tems and combining them with the phrase- based ones. The joint setups yield em- pirical gains of up to 1.6 points in BLEU and 1.0 points in TER on the WMT news- test2013 test set compared to the best sin- gle systems. 1 Introduction EU-BRIDGE1 is a European research project which is aimed at developing innovative speech translation technology. This paper describes a 1http://www.eu-bridge.eu joint WMT submission of three EU-BRIDGE project partners. RWTH Aachen University (RWTH), the University of Edinburgh (UEDIN) and Karlsruhe Institute of Technology (KIT) all provided several individual systems which were combined by means of the RWTH Aachen system combination approach (Freitag et al., 2014). As distinguished from our EU-BRIDGE joint submis- sion to the IWSLT 2013 evaluation campaign (Fre- itag et al., 2013), we particularly focused on trans- lation of news text (instead of talks) for WMT. Be- sides, we put an emphasis on engineering syntax- based systems in order to combine them with our more established phrase-based engines. We built combined system setups for translation from Ger- man to English as well as from English to Ger- man. This paper gives some insight into the tech- nology behind the system combination framework and the combined engines which have been used to produce the joint EU-BRIDGE submission to the WMT 2014 translation task. The remainder of the paper is structured as fol- lows: We first describe the individual systems by RWTH Aachen University (Section 2), the Uni- versity of Edinburgh (Section 3), and Karlsruhe Institute of Technology (Section 4). We then present the techniques for machine translation sys- tem combination in Section 5. Experimental re- sults are given in Section 6. We finally conclude the paper with Section 7. 2 RWTH Aachen University RWTH (Peitz et al., 2014) employs both the phrase-based (RWTH scss) and the hierarchical (RWTH hiero) decoder implemented in …

Content truncated to 500 words