DFKI’s experimental hybrid MT system for WMT 2015

Proceedings of the Tenth Workshop on Statistical Machine Translation, pages 66–73, Lisboa, Portugal, 17-18 September 2015. c⃝2015 Association for Computational Linguistics. DFKI’s experimental hybrid MT system for WMT 2015 Eleftherios Avramidis, Maja Popovi´c∗ and Aljoscha Burchardt German Research Center for Artificial intelligence (DFKI) Language Technology Lab firstname.lastname@dfki.de ∗ Humboldt University of Berlin maja.popovic@hu-berlin.de Abstract DFKI participated in the shared transla- tion task of WMT 2015 with the German- English language pair in each translation direction. The submissions were gener- ated using an experimental hybrid system based on three systems: a statistical Moses system, a commercial rule-based system, and a serial coupling of the two where the output of the rule-based system is fur- ther translated by Moses trained on paral- lel text consisting of the rule-based output and the original target language. The out- puts of three systems are combined using two methods: (a) an empirical selection mechanism based on grammatical features (primary submission) and (b) IBM1 mod- els based on POS 4-grams (contrastive sub- mission). 1 Introduction The system architecture we will describe has been developed within the QTLEAP project.1 The goal of the project is to explore different combinations of shallow and deep processing for improving MT quality. The system presented in this paper is the first of a series of MT system prototypes developed in the project. Figure 1 shows the overall architec- ture that includes: • A statistical Moses system, • the commercial transfer-based system Lucy, • their serial combination (”LucyMoses”), and • an informed selection mechanism (”ranker”). The components of this hybrid system will be detailed in the sections below. 1http://qtleap.eu/ Figure 1: Architecture of System Combination. 2 Translation systems Moses Our statistical machine translation system was based on a vanilla phrase-based system built with Moses (Koehn et al., 2007) trained on the cor- pora Europarl ver. 7, News Commentary ver. 9 (Bojar et al., 2014), Commoncrawl (Smith et al., 2013) and MultiUN (Eisele and Chen, 2010). Lan- guage models of order 5 have been built and inter- polated with SRILM (Stolcke, 2002) and KenLM (Heafield, 2011). For German to English, we also experimented with the method of pre-ordering the source side based on the target-side grammar (Popovi´c and Ney, 2006). As a tuning set we used the news-test 2013. Lucy The transfer-based Lucy system (Alonso and Thurmair, 2003) includes the results of long lin- guistic efforts over the last decades and that has been used in previous projects including EURO- MATRIX, EUROMATRIX+ and QTLAUNCHPAD, while relevant hybrid systems have been submit- ted to WMT (Chen et al., 2007; Federmann et al., 2010; Hunsicker et al., 2012). The transfer- based approach has shown good results that com- pete with pure statistical systems, whereas it fo- cuses on translating according to linguistic struc- 66 tures. Its functionality is based on hand-written linguistic rules and there are no major empirical components. Translations are processed on three phases: • the analysis phase, where the source- language text is parsed and a tree of the source language is constructed …

Content truncated to 500 words