Hallucinating Phrase Translations for Low Resource MT

We demonstrate that “hallucinating” phrasal translations can significantly improve the quality of machine translation in low resource conditions. Our hallucinated phrase tables consist of entries composed from multiple unigram translations drawn from the baseline phrase table and from translations that are induced from monolingual corpora. The hallucinated phrase table is very noisy. Its translations are low precision but high recall. We counter this by introducing 30 new feature functions (including a variety of monolinguallyestimated features) and by aggressively pruning the phrase table. Our analysis evaluates the intrinsic quality of our hallucinated phrase pairs as well as their impact in end-to-end Spanish-English and Hindi-English MT.

• Publication year

  2014

• Venue

  Conference on Computational Natural Language Learning

• Publication date

  2014-06-01

• Fields of study

  Linguistics, Computer Science

• Identifiers
  DOI 10.3115/v1/W14-1617
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Combining Bilingual and Comparable Corpora for Low Resource Machine Translation

  2013influential reference
• Dependency-Based Decipherment for Resource-Limited Machine Translation

  2013influential reference
• Graph Propagation for Paraphrasing Out-of-Vocabulary Words in Statistical Machine Translation

  2013cited by this paper
• Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation

  2013cited by this paper
• Supervised Bilingual Lexicon Induction with Multiple Monolingual Signals

  2013influential reference
• Scalable Decipherment for Machine Translation via Hash Sampling

  2013cited by this paper
• A Systematic Comparison of Phrase Table Pruning Techniques

  2012cited by this paper
• Toward Statistical Machine Translation without Parallel Corpora

  2012cited by this paper
• Prediction of Learning Curves in Machine Translation

  2012cited by this paper
• Constructing Parallel Corpora for Six Indian Languages via Crowdsourcing

  2012influential reference
• Large Scale Decipherment for Out-of-Domain Machine Translation

  2012cited by this paper
• Batch Tuning Strategies for Statistical Machine Translation

  2012cited by this paper
• Better Hypothesis Testing for Statistical Machine Translation: Controlling for Optimizer Instability

  2011cited by this paper
• Domain Adaptation for Machine Translation by Mining Unseen Words

  2011cited by this paper
• Deciphering Foreign Language

  2011cited by this paper
• Findings of the 2010 Joint Workshop on Statistical Machine Translation and Metrics for Machine Translation

  2010influential reference
• Extracting Parallel Sentences from Comparable Corpora using Document Level Alignment

  2010cited by this paper
• Discarding monotone composed rule for hierarchical phrase-based statistical machine translation

  2009cited by this paper
• Translating Compounds by Learning Component Gloss Translation Models via Multiple Languages

  2008cited by this paper
• Phrase Table Training for Precision and Recall: What Makes a Good Phrase and a Good Phrase Pair?

  2008cited by this paper
• Learning Bilingual Lexicons from Monolingual Corpora

  2008influential reference
• Moses: Open Source Toolkit for Statistical Machine Translation

  2007cited by this paper
• Improving Translation Quality by Discarding Most of the Phrasetable

  2007cited by this paper
• Extracting Parallel Sub-Sentential Fragments from Non-Parallel Corpora

  2006cited by this paper
• Europarl: A Parallel Corpus for Statistical Machine Translation

  2005cited by this paper
• A Systematic Comparison of Various Statistical Alignment Models

  2003cited by this paper
• Statistical Phrase-Based Translation

  2003cited by this paper
• Learning a Translation Lexicon from Monolingual Corpora

  2002cited by this paper
• Inducing Translation Lexicons via Diverse Similarity Measures and Bridge Languages

  2002cited by this paper
• Bleu: a Method for Automatic Evaluation of Machine Translation

  2002cited by this paper
• Identifying Word Translations in Non-Parallel Texts

  1995cited by this paper

• An Audit on the Perspectives and Challenges of Hallucinations in NLP

  2024cites this paper
• LexicoMatic: Automatic Creation of Multilingual Lexical-Semantic Dictionaries

  2023cites this paper
• Working Towards Digital Documentation of Uralic Languages With Open-Source Tools and Modern NLP Methods

  2023cites this paper
• Low-resource Neural Machine Translation: Methods and Trends

  2022cites this paper
• High-Resource Methodological Bias in Low-Resource Investigations

  2022cites this paper
• Organising lexica into analogical grids: a study of a holistic approach for morphological generation under various sizes of data in various languages

  2022cites this paper
• Documentación de lenguas amenazadas en la época digital

  2021cites this paper
• Cross-lingual Supervision Improves Unsupervised Neural Machine Translation

  2020cites this paper
• Multi-Agent Cross-Translated Diversification for Unsupervised Machine Translation

  2020cites this paper
• Cross-model Back-translated Distillation for Unsupervised Machine Translation

  2020cites this paper
• Unsupervised Neural Machine Translation with Indirect Supervision

  2020cites this paper
• Quasi-Parallel Corpora: Hallucinating Translations for the Chinese-Japanese Language Pair

  2018cites this paper
• Phrase Table Induction Using Monolingual Data for Low-Resource Statistical Machine Translation

  2018influential citation
• Phrase-Based & Neural Unsupervised Machine Translation

  2018cites this paper
• Integrated Speech and Language Technology for Intelligence, Surveillance, and Reconnaissance (ISR)

  2017cites this paper
• What ’ s in a Domain ? Analyzing Genre and Topic Differences in SMT

  2017cites this paper
• Phrase Table Induction Using In-Domain Monolingual Data for Domain Adaptation in Statistical Machine Translation

  2017cites this paper
• Enriching low resource Statistical Machine Translation using induced bilingual lexicons

  2017cites this paper
• Five Shades of Noise: Analyzing Machine Translation Errors in User-Generated Text

  2015cites this paper
• What’s in a Domain? Analyzing Genre and Topic Differences in Statistical Machine Translation

  2015cites this paper
• Paraphrases for Statistical Machine Translation

  2015influential citation
• Japanese news simplification: tak design, data set construction, and analysis of simplified text

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

  2015influential citation
• Using Comparable Corpora to Augment Statistical Machine Translation Models in Low Resource Settings

  2014cites this paper
• Vector Space Models for Phrase-based Machine Translation

  2014cites this paper