A Joint Dependency Model of Morphological and Syntactic Structure for Statistical Machine Translation
By Rico Sennrich, Barry Haddow. 2015
13 citations recorded by the source corpus.
When translating between two languages that differ in their degree of morphological synthesis, syntactic structures in one language may be realized as morphological structures in the other, and SMT models need a mechanism to learn such translations. Prior work has used morpheme splitting with flat representations that do not encode the hierarchical structure between morphemes, but this structure is relevant for learning morphosyntactic constraints and selectional preferences. We propose to model syntactic and morphological structure jointly in a dependency translation model, allowing the system to generalize to the level of morphemes. We present a dependency representation of German compounds and particle verbs that results in improvements in translation quality of 1.4‐1.8 BLEU in the WMT English‐German translation task.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Rico Sennrich
- Philipp Koehn
- Barry Haddow
- Felix Stahlberg
- Ondrej Bojar
- Yonatan Belinkov
- Kyunghyun Cho
- Graham Neubig
Citations in the corpus
- How Grammatical is Character-level Neural Machine Translation? Assessing MT Quality with Contrastive Translation Pairs (2016, methodology): “A similar method of assessment has previously been used for monolingual language models (Sennrich and Haddow, 2015; Linzen et al., 2016), and we apply it to the task of machine translation.”
- Edinburgh’s Syntax-Based Systems at WMT 2015 (2014, methodology): “, 2013); treebank binarization (Sennrich and Haddow, 2015); particle verb restructuring (Sennrich and Haddow, 2015).”
- Merged bilingual trees based on Universal Dependencies in Machine Translation (2016, background): “Tree-based machine translation systems (Chiang et al., 2005; Dušek et al., 2012; Sennrich and Haddow, 2015) are alternatives to the leading phrasebased MT systems (Koehn et al.”
- Edinburgh’s Statistical Machine Translation Systems for WMT16 (2016, methodology): “Trees are transformed through binarization and a hierarchical representation of morphologically complex words (Sennrich and Haddow, 2015).”
- Does Multimodality Help Human and Machine for Translation and Image Captioning? (2016, methodology): “In order to generalize better over the compound structs in German, we trained and applied a compound splitter3 (Sennrich and Haddow, 2015) over the German vocabulary of training and validation sets.”