CUNI in WMT15: Chimera Strikes Again

Proceedings of the Tenth Workshop on Statistical Machine Translation, pages 79–83, Lisboa, Portugal, 17-18 September 2015. c⃝2015 Association for Computational Linguistics. CUNI in WMT15: Chimera Strikes Again Ondˇrej Bojar and Aleˇs Tamchyna Charles University in Prague, Faculty of Mathematics and Physics Institute of Formal and Applied Linguistics Malostransk´e n´amˇest´ı 25, Prague, Czech Republic surname@ufal.mff.cuni.cz Abstract This paper describes our WMT15 system submission for the translation task, a hy- brid system for English-to-Czech transla- tion. We repeat the successful setup from the previous two years. 1 Introduction CHIMERA (Bojar et al., 2013; Tamchyna et al., 2014) is our English-to-Czech MT system de- signed as a combination of three very different components: • TectoMT (Popel and ˇZabokrtsk´y, 2010), a deep-syntactic transfer-based system, • Moses (Koehn et al., 2007), where we use a factored phrase-based setup with large lan- guage models, • Depfix (Rosa et al., 2012), an automatic post- editing system, aimed at correcting mainly errors in morphological agreement but suc- cessful also in semantic corrections, esp. re- covery of lost negation. The overall setup as well as the details on each of the components have been described in the past. We nevertheless briefly review it here, to make the paper self-contained. This year, our submission mainly differed in the additional data we were able to collect. We thus evaluate how much do the additional data help in contrast with an identical setup using WMT15 training data only.1 For the manual evaluation in WMT15, we submitted the non-constrained sys- tem, and even the “constrained” setup might not qualify as such, since it is a system combination and both TectoMT and Depfix rely on handcrafted rules to some extent. 1http://www.statmt.org/wmt15/ translation-task.html In the following, we provide various details of the setup. We leave Depfix aside, since we simply applied it as a post-processing step and the rele- vant analysis of its rules was published previously (Bojar et al., 2013). 2 Chimera in WMT15 2.1 Factored Setup We use our established setup, translating from English word form in one translation step to the Czech word form and morphological tag. This al- lows us to use language models over morphologi- cal tags, see §2.5 below. Our word forms are in truecase, i.e. the words at sentence beginnings are lowercased, unless they are names. We rely on Czech and English lemma- tizers2 to select the true case. Otherwise, our setup is fairly standard. We do not use any models of reordering, relying on basic distortion penalty. 2.2 Our System Combination The first two components of CHIMERA, Tec- toMT (which appears in WMT evaluations as CU- TECTOMT) and Moses are independent MT sys- tems on their own. CHIMERA combines them in a way remotely similar to standard system combi- nation techniques (Matusov et al., 2008) and adds the third component, Depfix, for automatic correc- tion of some grammar and semantic errors. For clarity, we will use the abbreviation CH to refer to the basic Moses setup without CU-TECTOMT. CH refers to the first stage, where CU-TECTOMT has …

Content truncated to 500 words