Syntax-based Statistical Machine Translation Research Articles

Incorporating target language semantic roles into a string-to-tree translation model

The string-to-tree model is one of the most successful syntax-based statistical machine translation (SMT) models. It models the grammaticality of the output via target-side syntax. However, it does not use any semantic information and tends to produce translations containing semantic role confusions and error chunk sequences. In this paper, we propose two methods to use semantic roles to improve the performance of the string-to-tree translation model: (1) adding role labels in the syntax tree; (2) constructing a semantic role tree, and then incorporating the syntax information into it. We then perform string-to-tree machine translation using the newly generated trees. Our methods enable the system to train and choose better translation rules using semantic information. Our experiments showed significant improvements over the state-of-the-art string-to-tree translation system on both spoken and news corpora, and the two proposed methods surpass the phrase-based system on large-scale training data.

Combining translation memories and statistical machine translation using sparse features

The combination of translation memories (TMs) and statistical machine translation (SMT) has been demonstrated to be beneficial. In this paper, we present a combination approach which integrates TMs into SMT by using sparse features extracted at run-time during decoding. These features can be used on both phrase-based SMT and syntax-based SMT. We conducted experiments on a publicly available English---French data set and an English---Spanish industrial data set. Our experimental results show that these features significantly improve our phrase-based and syntax-based SMT baselines on both language pairs.

Composition Closure of Linear Extended Top-down Tree Transducers

Linear extended top-down tree transducers (or synchronous tree-substitution grammars) are popular formal models of tree transformations that are extensively used in syntax-based statistical machine translation. The expressive power of compositions of such transducers with and without regular look-ahead is investigated. In particular, the restrictions of ź-freeness, strictness, and nondeletion are considered. The composition hierarchy turns out to be finite for all ź-free (all rules consume input) variants of these transducers except for the nondeleting ź-free transducers. The least number of transducers needed for the full expressive power of arbitrary compositions is presented. In all remaining cases (incl. the nondeleting ź-free transducers) the composition hierarchy does not collapse.

The Power of Weighted Regularity-Preserving Multi Bottom-Up Tree Transducers

The expressive power of regularity-preserving [Formula: see text]-free weighted linear multi bottom-up tree transducers is investigated. These models have very attractive theoretical and algorithmic properties, but (especially in the weighted setting) their expressive power is not well understood. Despite the regularity-preserving restriction, their power still exceeds that of composition chains of [Formula: see text]-free weighted linear extended top-down tree transducers with regular look-ahead. The latter devices are a natural super-class of weighted synchronous tree substitution grammars, which are commonly used in syntax-based statistical machine translation. In particular, the linguistically motivated discontinuous transformation of topicalization can be modeled by such multi bottom-up tree transducers, whereas the mentioned composition chains cannot implement it. On the negative side, the inverse of topicalization cannot be implemented by any such multi bottom-up tree transducer, which confirms their bottom-up nature (and non-closure under inverses). An interesting, promising, and widely applicable proof technique is used to prove these statements.

A Substitution-Translation-Restoration Framework for Handling Unknown Words in Statistical Machine Translation

Unknown words are one of the key factors that greatly affect the translation quality. Traditionally, nearly all the related researches focus on obtaining the translation of the unknown words. However, these approaches have two disadvantages. On the one hand, they usually rely on many additional resources such as bilingual web data; on the other hand, they cannot guarantee good reordering and lexical selection of surrounding words. This paper gives a new perspective on handling unknown words in statistical machine translation (SMT). Instead of making great efforts to find the translation of unknown words, we focus on determining the semantic function of the unknown word in the test sentence and keeping the semantic function unchanged in the translation process. In this way, unknown words can help the phrase reordering and lexical selection of their surrounding words even though they still remain untranslated. In order to determine the semantic function of an unknown word, we employ the distributional semantic model and the bidirectional language model. Extensive experiments on both phrase-based and linguistically syntax-based SMT models in Chinese-to-English translation show that our method can substantially improve the translation quality.

Syntax-Based Post-Ordering for Efficient Japanese-to-English Translation

This article proposes a novel reordering method for efficient two-step Japanese-to-English statistical machine translation (SMT) that isolates reordering from SMT and solves it after lexical translation. This reordering problem, called post-ordering , is solved as an SMT problem from Head-Final English (HFE) to English. HFE is syntax-based reordered English that is very successfully used for reordering with English-to-Japanese SMT. The proposed method incorporates its advantage into the reverse direction, Japanese-to-English, and solves the post-ordering problem by accurate syntax-based SMT with target language syntax. Two-step SMT with the proposed post-ordering empirically reduces the decoding time of the accurate but slow syntax-based SMT by its good approximation using intermediate HFE. The proposed method improves the decoding speed of syntax-based SMT decoding by about six times with comparable translation accuracy in Japanese-to-English patent translation experiments.

Language Modeling for Syntax-Based Machine Translation Using Tree Substitution Grammars

The poor grammatical output of Machine Translation (MT) systems appeals syntax-based approaches within language modeling. However, previous studies showed that syntax-based language modeling using (Context-Free) Treebank Grammars was not very helpful in improving BLEU scores for Chinese-English machine translation. In this article we further study this issue in the context of Chinese-English syntax-based Statistical Machine Translation (SMT) where Synchronous Tree Substitution Grammars (STSGs) are utilized to model the translation process. In particular, we develop a Tree Substitution Grammar-based language model for syntax-based MT, and present three methods to efficiently integrate the proposed language model into MT decoding. In addition, we design a simple and effective method to adapt syntax-based language models for MT tasks. We demonstrate that the proposed methods are able to benefit a state-of-the-art syntax-based MT system. On the NIST Chinese-English MT evaluation corpora, we finally achieve an improvement of 0.6 BLEU points over the baseline.

Re-structuring, Re-labeling, and Re-aligning for Syntax-Based Machine Translation

This article shows that the structure of bilingual material from standard parsing and alignment tools is not optimal for training syntax-based statistical machine translation (SMT) systems. We present three modifications to the MT training data to improve the accuracy of a state-of-the-art syntax MT system: re-structuring changes the syntactic structure of training parse trees to enable reuse of substructures; re-labeling alters bracket labels to enrich rule application context; and re-aligning unifies word alignment across sentences to remove bad word alignments and refine good ones. Better structures, labels, and word alignments are learned by the EM algorithm. We show that each individual technique leads to improvement as measured by BLEU, and we also show that the greatest improvement is achieved by combining them. We report an overall 1.48 BLEU improvement on the NIST08 evaluation set over a strong baseline in Chinese/English translation.