We present a perceptron-style discriminative approach to machine translation in which large feature sets can be exploited. Unlike discriminative reranking approaches, our system can take advantage of learned features in all stages of decoding. We first discuss several challenges to error-driven discriminative approaches. In particular, we explore different ways of updating parameters given a training example. We find that making frequent but smaller updates is preferable to making fewer but larger updates. Then, we discuss an array of features and show both how they quantitatively increase BLEU score and how they qualitatively interact on specific examples. One particular feature we investigate is a novel way to introduce learning into the initial phrase extraction process, which has previously been entirely heuristic. 1

We argue that the machine translation community is overly reliant on the Bleu machine translation evaluation metric. We show that an improved Bleu score is neither necessary nor sufficient for achieving an actual improvement in translation quality, and give two significant counterexamples to Bleu’s correlation with human judgments of quality. This offers new potential for research which was previously deemed unpromising by an inability to improve upon Bleu scores. 1

We use the Margin Infused Relaxed Algorithm of Crammer et al. to add a large number of new features to two machine translation systems: the Hiero hierarchical phrasebased translation system and our syntax-based translation system. On a large-scale Chinese-English translation task, we obtain statistically significant improvements of +1.5 Bleu and +1.1 Bleu, respectively. We analyze the impact of the new features and the performance of the learning algorithm. 1

We investigate a number of simple methods for improving the word-alignment accuracy of IBM Model 1. We demonstrate reduction in alignment error rate of approximately 30 % resulting from (1) giving extra weight to the probability of alignment to the null word, (2) smoothing probability estimates for rare words, and (3) using a simple heuristic estimation method to initialize, or replace, EM training of model parameters.

Automatic evaluation of machine translation, based on computing n-gram similarity between system output and human reference translations, has revolutionized the development of MT systems. We explore the use of syntactic information, including constituent labels and head-modifier dependencies, in computing similarity between output and reference. Our results show that adding syntactic information to the evaluation metric improves both sentence-level and corpus-level correlation with human judgments.

Statistical machine translation (SMT) treats the translation of natural language as a machine learning problem. By examining many samples of human-produced translation, SMT algorithms automatically learn how to translate. SMT has made tremendous strides in less than two decades, and many popular techniques have only emerged within the last few years. This survey presents a tutorial overview of state-of-the-art SMT at the beginning of 2007. We begin with the context of the current research, and then move to a formal problem description and an overview of the four main subproblems: translational equivalence modeling, mathematical modeling, parameter estimation, and decoding. Along the way, we present a taxonomy of some different approaches within these areas. We conclude with an overview of evaluation and notes on future directions.

done while at Microsoft Research We describe a novel approach to statistical machine translation that combines syntactic information in the source language with recent advances in phrasal translation. We depend on a source-language dependency parser and a word-aligned parallel corpus. The only target language resource assumed is a word breaker. These are used to produce treelet (“phrase”) translation pairs as well as several models, including a channel model, an order model, and a target language model. Together these models and the treelet translation pairs provide a powerful and promising approach to MT that incorporates the power of phrasal SMT with the linguistic generality available in a parser. We evaluate two decoding approaches, one inspired by dynamic programming and the

While phrase-based statistical machine translation systems currently deliver state-of-theart performance, they remain weak on word order changes. Current phrase reordering models can properly handle swaps between adjacent phrases, but they typically lack the ability to perform the kind of long-distance reorderings possible with syntax-based systems. In this paper, we present a novel hierarchical phrase reordering model aimed at improving non-local reorderings, which seamlessly integrates with a standard phrase-based system with little loss of computational efficiency. We show that this model can successfully handle the key examples often used to motivate syntax-based systems, such as the rotation of a prepositional phrase around a noun phrase. We contrast our model with reordering models commonly used in phrase-based systems, and show that our approach provides statistically significant BLEU point gains for two language pairs: Chinese-English (+0.53 on MT05 and +0.71 on MT08) and Arabic-English (+0.55 on MT05). 1

We investigate some pitfalls regarding the discriminatory power of MT evaluation metrics and the accuracy of statistical significance tests. In a discriminative reranking experiment for phrase-based SMT we show that the NIST metric is more sensitive than BLEU or F-score despite their incorporation of aspects of fluency or meaning adequacy into MT evaluation. In an experimental comparison of two statistical significance tests we show that p-values are estimated more conservatively by approximate randomization than by bootstrap tests, thus increasing the likelihood of type-I error for the latter. We point out a pitfall of randomly assessing significance in multiple pairwise comparisons, and conclude with a recommendation to combine NIST with approximate randomization, at more stringent rejection levels than is currently standard.

Hierarchical organization is a well known property of language, and yet the notion of hierarchical structure has been largely absent from the best performing machine translation systems in recent community-wide evaluations. In this paper, we discuss a new hierarchical phrase-based statistical machine translation system (Chiang, 2005), presenting recent extensions to the original proposal, new evaluation results in a community-wide evaluation, and a novel technique for fine-grained comparative analysis of MT systems. 1