Lattice-based Minimum Error Rate Training for Statistical Machine Translation

Minimum Error Rate Training (MERT) is an effective means to estimate the feature function weights of a linear model such that an automated evaluation criterion for measuring system performance can directly be optimized in training. To accomplish this, the training procedure determines for each feature function its exact error surface on a given set of candidate translations. The feature function weights are then adjusted by traversing the error surface combined over all sentences and picking those values for which the resulting error count reaches a minimum. Typically, candidates in MERT are represented as N-best lists which contain the N most probable translation hypotheses produced by a decoder. In this paper, we present a novel algorithm that allows for efficiently constructing and representing the exact error surface of all translations that are encoded in a phrase lattice. Compared to N-best MERT, the number of candidate translations thus taken into account increases by several orders of magnitudes. The proposed method is used to train the feature function weights of a phrase-based statistical machine translation system. Experiments conducted on the NIST 2008 translation tasks show significant runtime improvements and moderate BLEU score gains over N-best MERT.

• Publication year

  2008

• Venue

  Conference on Empirical Methods in Natural Language Processing

• Publication date

  2008-10-25

• Fields of study

  Computer Science

• Identifiers
  DOI 10.3115/1613715.1613807
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Beyond Log-Linear Models: Boosted Minimum Error Rate Training for N-best Re-ranking

  2008cited by this paper
• Lattice Minimum Bayes-Risk Decoding for Statistical Machine Translation

  2008cited by this paper
• Regularization and Search for Minimum Error Rate Training

  2008cited by this paper
• An Empirical Study on Computing Consensus Translations from Multiple Machine Translation Systems

  2007cited by this paper
• Numerical Recipes 3rd Edition: The Art of Scientific Computing

  2007cited by this paper
• Online Large-Margin Training for Statistical Machine Translation

  2007cited by this paper
• A Systematic Comparison of Training Criteria for Statistical Machine Translation

  2007cited by this paper
• Combining Outputs from Multiple Machine Translation Systems

  2007cited by this paper
• Minimum Risk Annealing for Training Log-Linear Models

  2006cited by this paper
• BOUNDS ON THE NUMBER OF INFERENCE FUNCTIONS OF A GRAPHICAL MODEL

  2006cited by this paper
• Syntax Augmented Machine Translation via Chart Parsing

  2006cited by this paper
• A Hierarchical Phrase-Based Model for Statistical Machine Translation

  2005cited by this paper
• Minimum Bayes-Risk Decoding for Statistical Machine Translation

  2004cited by this paper
• Minimum Error Rate Training in Statistical Machine Translation

  2003cited by this paper
• Bleu: a Method for Automatic Evaluation of Machine Translation

  2002cited by this paper
• Generation of Word Graphs in Statistical Machine Translation

  2002cited by this paper
• Discriminative training via linear programming

  1999cited by this paper
• Numerical Recipes in C The Art of Scientific Computing

  1995cited by this paper
• Numerical Recipes: The Art of Scientific Computing

  1989cited by this paper
• Numerical recipes: The art of scientific computing

  1987cited by this paper
• Algorithms for Reporting and Counting Geometric Intersections

  1979cited by this paper

• Variational Prefix Tuning for Diverse and Accurate Code Summarization Using Pre-trained Language Models

  2025cites this paper
• The Roles of Language Models and Hierarchical Models in Neural Sequence-to-Sequence Prediction

  2020influential citation
• Advances in Computational Linguistics

  2020cites this paper
• A simple discriminative training method for machine translation with large-scale features

  2019cites this paper
• Perturbation Based Learning for Structured NLP Tasks with Application to Dependency Parsing

  2019cites this paper
• Efficient Phrase Table pruning for Hindi to English machine translation through syntactic and marker-based filtering and hybrid similarity measurement

  2018cites this paper
• Incorporating Chinese Radicals Into Neural Machine Translation: Deeper Than Character Level

  2018cites this paper
• Statistical Machine Translation of English Text to API Code Usages :

  2018cites this paper
• Phrase-Based Statistical Model for Korean Morpheme Segmentation and POS Tagging

  2018influential citation
• The University of Cambridge’s Machine Translation Systems for WMT18

  2018cites this paper
• Apply Chinese Radicals Into Neural Machine Translation: Deeper Than Character Level

  2018cites this paper
• Preference learning for machine translation

  2018cites this paper
• 18 Algorithms for MT 2 : Parameter Optimization Methods

  2018cites this paper
• Latent domain models for statistical machine translation

  2017cites this paper
• Extensions to the PRESEMT Methodology

  2017cites this paper
• A survey of domain adaptation for statistical machine translation

  2017cites this paper
• Multi-Language Online Handwriting Recognition

  2017cites this paper
• ICDAR2017 Competition on Recognition of Early Indian Printed Documents - REID2017

  2017cites this paper
• Hierarchical statistical semantic translation and realization

  2017influential citation
• Source sentence simplification for statistical machine translation

  2017cites this paper
• Models and Inference for Prefix-Constrained Machine Translation

  2016cites this paper
• The Edit Distance Transducer in Action: The University of Cambridge English-German System at WMT16

  2016cites this paper
• Bilingual Language Models using Word Embeddings for Machine Translation

  2016cites this paper
• Data-driven natural language generation using statistical machine translation and discriminative learning. (L'approche discriminante à la génération de la parole)

  2016cites this paper
• Phrase-based Machine Translation using Multiple Preordering Candidates

  2016cites this paper
• An Attentional Neural Conversation Model with Improved Specificity

  2016cites this paper
• Integrating Morphological Desegmentation into Phrase-based Decoding

  2016cites this paper
• Optimization for Statistical Machine Translation: A Survey

  2016cites this paper
• Syntactically Guided Neural Machine Translation

  2016cites this paper
• A Study of Statistical Machine Translation Methods for Under Resourced Languages

  2016cites this paper
• Comparison of Grapheme-to-Phoneme Conversion Methods on a Myanmar Pronunciation Dictionary

  2016cites this paper
• A new model for persian multi-part words edition based on statistical machine translation

  2016cites this paper
• The 54th Annual Meeting of the Association for Computational Linguistics

  2016cites this paper
• Interlocking Phrases in Phrase-based Statistical Machine Translation

  2016cites this paper
• Neural Machine Translation by Minimising the Bayes-risk with Respect to Syntactic Translation Lattices

  2016cites this paper
• Domain adaptation for statistical machine translation

  2016cites this paper
• A Simple, Fast Diverse Decoding Algorithm for Neural Generation

  2016influential citation
• Mutual Information and Diverse Decoding Improve Neural Machine Translation

  2016cites this paper
• Pivot-based Statistical Machine Translation for Morphologically Rich Languages

  2016cites this paper
• The Application of Machine Transliteration Techniques to Spelling Correction

  2015cites this paper
• Transducer Disambiguation with Sparse Topological Features

  2015cites this paper
• Hierarchical Statistical Semantic Realization for Minimal Recursion Semantics

  2015cites this paper
• The Edinburgh machine translation systems for IWSLT 2015

  2015cites this paper
• Improving Statistical Machine Translation with a Multilingual Paraphrase Database

  2015cites this paper
• A Large-scale Study of Statistical Machine Translation Methods for Khmer Language

  2015cites this paper
• Paraphrases for Statistical Machine Translation

  2015cites this paper
• The Geometry of Statistical Machine Translation

  2015influential citation
• Label transition and selection pruning and automatic decoding parameter optimization for time-synchronous Viterbi decoding

  2015cites this paper
• Learning to Rank Algorithms and Their Application in Machine Translation

  2015cites this paper
• English to Japanese spoken lecture translation system by using DNN-HMM and phrase-based SMT

  2015cites this paper
• Efficient Top-Down BTG Parsing for Machine Translation Preordering

  2015cites this paper
• Hierarchical Phrase-based Stream Decoding

  2015cites this paper
• The role of syntax and semantics in machine translation and quality estimation of machine-translated user-generated content

  2015cites this paper
• What Matters Most in Morphologically Segmented SMT Models?

  2015cites this paper
• A Synchronous Hyperedge Replacement Grammar based approach for AMR parsing

  2015cites this paper
• Une méthode discriminant formation simple pour la traduction automatique avec Grands Caractéristiques

  2015cites this paper
• Neural Network Transduction Models in Transliteration Generation

  2015cites this paper
• Mixed domain vs. multi-domain statistical machine translation

  2015cites this paper
• Improving translation quality stability using Bayesian predictive adaptation

  2015cites this paper
• English to Japanese spoken language translation system for classroom lectures

  2014cites this paper
• Multimodal interactive structured prediction.

  2014cites this paper
• Sampling Tree Fragments from Forests

  2014cites this paper
• Effective Incorporation of Source Syntax into Hierarchical Phrase-based Translation

  2014cites this paper
• Unlabeled Dependency Parsing Based Pre-reordering for Chinese-to-Japanese SMT

  2014cites this paper
• Learning from Post-Editing: Online Model Adaptation for Statistical Machine Translation

  2014influential citation
• An Empirical Comparison of Features and Tuning for Phrase-based Machine Translation

  2014cites this paper
• Lattice Desegmentation for Statistical Machine Translation

  2014cites this paper
• Online optimisation of log-linear weights in interactive machine translation

  2014cites this paper
• Phrase-Based Statistical Translation of Programming Languages

  2014cites this paper
• Morphological Analysis for Japanese Noisy Text based on Character-level and Word-level Normalization

  2014cites this paper
• Improving Word Alignment Using Linguistic Code Switching Data

  2014cites this paper
• Human Effort and Machine Learnability in Computer Aided Translation

  2014cites this paper
• Handling OOV Words in Dialectal Arabic to English Machine Translation

  2014cites this paper
• Phrasal: A Toolkit for New Directions in Statistical Machine Translation

  2014cites this paper
• Coordinate System Selection for Minimum Error Rate Training in Statistical Machine Translation

  2014cites this paper
• Comparison of data selection techniques for the translation of video lectures

  2014cites this paper
• Query Lattice for Translation Retrieval

  2014cites this paper
• Comparing Representations of Semantic Roles for String-To-Tree Decoding

  2014cites this paper
• Bayesian Optimisation for Machine Translation

  2014influential citation
• Sentence Level Dialect Identification for Machine Translation System Selection

  2014cites this paper
• The usage of differential evolution in a statistical machine translation

  2014cites this paper
• Unsupervised Quality Estimation Model for English to German Translation and Its Application in Extensive Supervised Evaluation

  2014cites this paper
• The Multilingual Paraphrase Database

  2014cites this paper
• Jane: Open Source Machine Translation System Combination

  2014cites this paper
• An Improved Differential Evolution Algorithm Based on Statistical Log-linear Model

  2013cites this paper
• LEVERAGING DIVERSE SOURCES IN STATISTICAL MACHINE TRANSLATION

  2013cites this paper
• Source-Side Discontinuous Phrases for Machine Translation: A Comparative Study on Phrase Extraction and Search

  2013cites this paper
• Lattice BLEU oracles in machine translation

  2013cites this paper
• A Systematic Exploration of Diversity in Machine Translation

  2013cites this paper
• Multiword Expressions in the Context of Statistical Machine Translation

  2013cites this paper
• Improved Online Learning and Modeling for Feature-Rich Discriminative Machine Translation

  2013cites this paper
• Amélioration des systèmes de traduction par analyse linguistique et thématique : application à la traduction depuis l'arabe. (Improvements for Machine Translation Systems Using Linguistic and Thematic Analysis : an Application to the Translation from Arabic)

  2013cites this paper
• Ensemble Triangulation for Statistical Machine Translation

  2013cites this paper
• Machine Translation for Human Translators

  2013cites this paper
• Dialectal Arabic to English Machine Translation: Pivoting through Modern Standard Arabic

  2013cites this paper
• Regularized Minimum Error Rate Training

  2013influential citation
• Morphological Analysis and Disambiguation for Dialectal Arabic

  2013cites this paper
• A Purely Monotonic Approach to Machine Translation for Similar Languages

  2013cites this paper
• Joint Language and Translation Modeling with Recurrent Neural Networks

  2013cites this paper
• Stacking for Statistical Machine Translation

  2013cites this paper