We present and compare various methods for computing word alignments using statistical or heuristic models. We consider the five alignment models presented in Brown, Della Pietra, Della Pietra, and Mercer (1993), the hidden Markov alignment model, smoothing techniques, and refinements. These statistical models are compared with two heuristic models based on the Dice coefficient. We present different methods for combining word alignments to perform a symmetrization of directed statistical alignment models. As evaluation criterion, we use the quality of the resulting Viterbi alignment compared to a manually produced reference alignment. We evaluate the models on the German-English Verbmobil task and the French-English Hansards task. We perform a detailed analysis of various design decisions of our statistical alignment system and evaluate these on training corpora of various sizes. An important result is that refined alignment models with a first-order dependence and a fertility model yield significantly better results than simple heuristic models. In the Appendix, we present an efficient training algorithm for the alignment models presented.

/pdf/a-systematic-comparison-of-various-statistical-alignment-3pcl3i9uz3.pdf

A systematic comparison of various statistical alignment models

In this paper, we describe improved alignment models for statistical machine translation. The statistical translation approach uses two types of information: a translation model and a language model. The language model used is a bigram or general m-gram model. The translation model is decomposed into a lexical and an alignment model. We describe two different approaches for statistical translation and present experimental results. The first approach is based on dependencies between single words, the second approach explicitly takes shallow phrase structures into account, using two different alignment levels: a phrase level alignment between phrases and a word level alignment between single words. We present results using the Verbmobil task (German-English, 6000word vocabulary) which is a limited-domain spoken-language task. The experimental tests were performed on both the text transcription and the speech recognizer output. 1 S t a t i s t i c a l M a c h i n e T r a n s l a t i o n The goal of machine translation is the translation of a text given in some source language into a target language. We are given a source string f / = fl...fj...fJ, which is to be translated into a target string e{ = el...ei...ex. Among all possible target strings, we will choose the string with the highest probability: = argmax {Pr(ezIlflJ)}

Improved Alignment Models for Statistical Machine Translation

We present an extension of phrase-based statistical machine translation models that enables the straight-forward integration of additional annotation at the word-level — may it be linguistic markup or automatically generated word classes. In a number of experiments we show that factored translation models lead to better translation performance, both in terms of automatic scores, as well as more grammatical coherence.

https://homepages.inf.ed.ac.uk/pkoehn/publications/emnlp2007-factored.pdf

Factored Translation Models

A cooperative conversational voice user interface is provided. The cooperative conversational voice user interface may build upon short-term and long-term shared knowledge to generate one or more explicit and/or implicit hypotheses about an intent of a user utterance. The hypotheses may be ranked based on varying degrees of certainty, and an adaptive response may be generated for the user. Responses may be worded based on the degrees of certainty and to frame an appropriate domain for a subsequent utterance. In one implementation, misrecognitions may be tolerated, and conversational course may be corrected based on subsequent utterances and/or responses.

System and method for a cooperative conversational voice user interface

We present a simple method for transferring dependency parsers from source languages with labeled training data to target languages without labeled training data. We first demonstrate that delexicalized parsers can be directly transferred between languages, producing significantly higher accuracies than unsupervised parsers. We then use a constraint driven learning algorithm where constraints are drawn from parallel corpora to project the final parser. Unlike previous work on projecting syntactic resources, we show that simple methods for introducing multiple source languages can significantly improve the overall quality of the resulting parsers. The projected parsers from our system result in state-of-the-art performance when compared to previously studied unsupervised and projected parsing systems across eight different languages.

/pdf/multi-source-transfer-of-delexicalized-dependency-parsers-zp99s8yy8y.pdf

Multi-Source Transfer of Delexicalized Dependency Parsers

Machine translation of a source language sentence involves selecting appropriate target language words and ordering the selected words to form a well-formed target language sentence. Most of the previous work on statistical machine translation relies on (local) associations of target words/phrases with source words/phrases for lexical selection. In contrast, in this paper, we present a novel approach to lexical selection where the target words are associated with the entire source sentence (global) without the need to compute local associations. Further, we present a technique for reconstructing the target language sentence from the selected words. We compare the results of this approach against those obtained from a finite-state based statistical machine translation system which relies on local lexical associations.

Statistical Machine Translation through Global Lexical Selection and Sentence Reconstruction

We describe a method for the fully automatic learning of hierarchical finite state translation models. The input to the method is transcribed speech utterances and their corresponding human translations, and the output is a set of head transducers, i.e. statistical lexical head-outward transducers. A word-alignment function and a head-ranking function are first obtained, and then counts are generated for hypothesized state transitions of head transducers whose lexical translations and word order changes are consistent with the alignment. The method has been applied to create an English-Spanish translation model for a Speech translation application, with word accuracy of over 75% as measured by a string-distance comparison to three reference translations.

https://faculty.washington.edu/fxia/courses/LING580MT/alshawi_ACL1998.pdf

Automatic Acquisition of Hierarchical Transduction Models for Machine Translation

Disclosed are systems, methods, and computer-readable media for performing translations from a source language to a target language The method comprises receiving a source phrase, generating a target bag of words based on a global lexical selection of words that loosely couples the source words/phrases and target words/phrases, and reconstructing a target phrase or sentence by considering all permutations of words with a conditional probability greater than a threshold

Machine Translation Using Global Lexical Selection and Sentence Reconstruction

Statistical phrase-based machine translation models crucially rely on word alignments. The search for word-alignments assumes a model of word locality between source and target languages that is violated in starkly different word-order languages such as English-Hindi. In this article, we present models that decouple the steps of lexical selection and lexical reordering with the aim of minimizing the role of word-alignment in machine translation. Indian languages are morphologically rich and have relatively free-word order where the grammatical role of content words is largely determined by their case markers and not just by their positions in the sentence. Hence, lexical selection plays a far greater role than lexical reordering. For lexical selection, we investigate models that take the entire source sentence into account and evaluate their performance for English-Hindi translation in a tourism domain.

Discriminative Machine Translation Using Global Lexical Selection

Machine translation of a source language sentence involves selecting appropriate target language words and ordering the selected words to form a well-formed target language sentence. Most of the previous work on statistical machine translation relies on (local) associations of target words/phrases with source words/phrases for lexical selection. In contrast, in this paper, we present a novel approach to lexical selection where the target words are associated with the entire source sentence (global) without the need for local associations. This technique is used by three models (Bag-of-words model, sequential model and hierarchical model) which predict the target language words given a source sentence and then order the words appropriately. We show that a hierarchical model performs best when compared to the other two models.

Srinivas Bangalore

Papers

Statistical Machine Translation through Global Lexical Selection and Sentence Reconstruction

Automatic Acquisition of Hierarchical Transduction Models for Machine Translation

Machine Translation Using Global Lexical Selection and Sentence Reconstruction

Discriminative Machine Translation Using Global Lexical Selection

Three models for discriminative machine translation using global lexical selection and sentence reconstruction