A discriminative matching approach to word alignment

Taskar, Ben; Lacoste-Julien, Simon; Klein, Dan

doi:10.3115/1220575.1220585

Cited by 72 publications

(114 citation statements)

References 17 publications

Supporting

Mentioning

113

Contrasting

Order By: Relevance

“…For example, various discriminative models have been proposed in the literature (Liu et al, 2005;Moore, 2005;Taskar et al, 2005). Their advantage is that they may integrate a wide range of features that may lead to improved alignment quality.…”

Section: Word Alignment and Smtmentioning

confidence: 99%

Estimating Word Alignment Quality for SMT Reordering Tasks

Stymne¹,

Tiedemann²,

Nivre³

2014

Proceedings of the Ninth Workshop on Statistical Machine Translation

View full text Add to dashboard Cite

Previous studies of the effect of word alignment on translation quality in SMT generally explore link level metrics only and mostly do not show any clear connections between alignment and SMT quality. In this paper, we specifically investigate the impact of word alignment on two pre-reordering tasks in translation, using a wider range of quality indicators than previously done. Experiments on German-English translation show that reordering may require alignment models different from those used by the core translation system. Sparse alignments with high precision on the link level, for translation units, and on the subset of crossing links, like intersected HMM models, are preferred. Unlike SMT performance the desired alignment characteristics are similar for small and large training data for the pre-reordering tasks. Moreover, we confirm previous research showing that the fuzzy reordering score is a useful and cheap proxy for performance on SMT reordering tasks.

show abstract

Section: Word Alignment and Smtmentioning

confidence: 99%

Estimating Word Alignment Quality for SMT Reordering Tasks

Stymne¹,

Tiedemann²,

Nivre³

2014

Proceedings of the Ninth Workshop on Statistical Machine Translation

View full text Add to dashboard Cite

show abstract

“…The class of at most 1-to-1 alignment matchings, A 1-1 , has been considered in several works (Melamed, 2000;Taskar et al, 2005;Moore et al, 2006). The alignment that maximizes a set of potentials factored as in Equation (1) can be found in O(n 3 ) time using a bipartite matching algorithm (Kuhn, 1955).…”

Section: -To-1 Matchingsmentioning

confidence: 99%

“…Initially, as in Taskar et al (2005) and Moore et al (2006), we assume the score a of a potential alignment a) decomposes as…”

Section: Alignment Familiesmentioning

confidence: 99%

See 1 more Smart Citation

Better word alignments with supervised ITG models

Haghighi

Blitzer

DeNero

et al. 2009

Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural

Self Cite

View full text Add to dashboard Cite

This work investigates supervised word alignment methods that exploit inversion transduction grammar (ITG) constraints. We consider maximum margin and conditional likelihood objectives, including the presentation of a new normal form grammar for canonicalizing derivations. Even for non-ITG sentence pairs, we show that it is possible learn ITG alignment models by simple relaxations of structured discriminative learning objectives. For efficiency, we describe a set of pruning techniques that together allow us to align sentences two orders of magnitude faster than naive bitext CKY parsing. Finally, we introduce many-to-one block alignment features, which significantly improve our ITG models. Altogether, our method results in the best reported AER numbers for Chinese-English and a performance improvement of 1.1 BLEU over GIZA++ alignments.

show abstract

“…More work on highly expressive models for parsing is given in [17], and the use of structural SVMs for protein threading is described in [10,12]. An alternative approach to alignment is [18]. Work on sequence tagging for natural language problems and OCR is given in [16,1].…”

Section: Application Examples and Related Workmentioning

confidence: 99%

Structured Output Prediction with Support Vector Machines

Joachims

2006

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. This abstract accompanying a presentation at S+SSPR 2006 explores the use of Support Vector Machines (SVMs) for predicting structured objects like trees, equivalence relations, or alignments. It is shown that SVMs can be extended to these problems in a well-founded way, still leading to a convex quadratic training problem and maintaining the ability to use kernels. While the training problem has exponential size, there is a simple algorithm that allows training in polynomial time. The algorithm is implemented in the SVM-Struct software, and it is discussed how the approach can be applied to problems ranging from natural language parsing to supervised clustering.

show abstract

A discriminative matching approach to word alignment

Cited by 72 publications

References 17 publications

Estimating Word Alignment Quality for SMT Reordering Tasks

Estimating Word Alignment Quality for SMT Reordering Tasks

Better word alignments with supervised ITG models

Structured Output Prediction with Support Vector Machines

Contact Info

Product

Resources

About