We present a method to transliterate names in the framework of end-to-end statistical machine translation. The system is trained to learn when to transliterate. For Arabic to English MT, we developed and trained a transliterator on a bitext of 7 million sentences and Google’s English terabyte ngrams and achieved better name translation accuracy than 3 out of 4 professional translators. The pape...

This paper studies a strategy for identifying and using multi-word expressions in Statistical Machine Translation. The performance of the proposed strategy for various types of multi-word expressions (like nouns or verbs) is evaluated in terms of alignment quality as well as translation accuracy. Evaluations are performed by using real-life data, namely the European Parliament corpus. Results f...

Syntax-based translation models that operate on the output of a source-language parser have been shown to perform better if allowed to choose from a set of possible parses. In this paper, we investigate whether this is because it allows the translation stage to overcome parser errors or to override the syntactic structure itself. We find that it is primarily the latter, but that under the right...

This paper reports the NTT statistical translation system participating in the evaluation campaign of IWSLT 2005. The NTT system is based on a phrase translation model and utilizes a large number of features with a log-linear model. We studied the various features recently developed in this research field and evaluate the system using supplied data as well as publicly available Chinese, Japanes...

A basic task in machine translation is to choose the right translation for source words with several possible translations in the target language. In this paper we treat word translation as a word sense disambiguation problem and train memory-based classifiers on words with alternative translations. The training data was automatically labeled with the corresponding translations by word-aligning...

The faithful and rapid translation of genetic information into peptide sequences is an indispensable property of the ribosome. The mechanistic understanding of strategies used by the ribosome to achieve both speed and fidelity during translation results from nearly a half century of biochemical and structural studies. Emerging from these studies is the common theme that the ribosome uses local ...

This paper presents modeling of transliteration as a phrase-based machine translation system. We used a popular phrasebased machine translation system for English-Hindi machine transliteration. We have achieved an accuracy of 38.1% on the test set. We used some basic rules to modulate the existing phrased-based transliteration system. Our experiments show that phrase-based machine translation s...

In this article we describe a method of translation from a deterministic Turing machine to an extended mH splicing system. The accuracy of this translation is proved by a theorem which shows that the only string present in single copy is a coding of a connguration of the simulated Turing machine. A simpler translation from a universal Turing machine with three tapes to an extended mH system is ...

In this paper, we improve the attention or alignment accuracy of neural machine translation by utilizing the alignments of training sentence pairs. We simply compute the distance between the machine attentions and the “true” alignments, and minimize this cost in the training procedure. Our experiments on large-scale Chinese-to-English task show that our model improves both translation and align...