Learning to Automatically Post-Edit Dropped Words in MT

Automatic post-editors (APEs) can improve adequacy of MT output by detecting and reinserting dropped content words, but the location where these words are inserted is critical. In this paper, we describe a probabilistic approach for learning reinsertion rules for specific languages and MT systems, as well as a method for synthesizing training data from reference translations. We test the insertion logic on MT systems for Chinese to English and Arabic to English. Our adaptive APE is able to insert within 3 words of the best location 73% of the time (32% in the exact location) in Arabic-English MT output, and 67% of the time in Chinese-English output (30% in the exact location), and delivers improved performance on automated adequacy metrics over a previous rule-based approach to insertion. We consider how particular aspects of the insertion problem make it particularly amenable to machine learning solutions.

[1]  Nizar Habash,et al.  MADA + TOKAN : A Toolkit for Arabic Tokenization , Diacritization , Morphological Disambiguation , POS Tagging , Stemming and Lemmatization , 2009 .

[2]  Philipp Koehn,et al.  Moses: Open Source Toolkit for Statistical Machine Translation , 2007, ACL.

[3]  Salim Roukos,et al.  Bleu: a Method for Automatic Evaluation of Machine Translation , 2002, ACL.

[4]  Salim Roukos,et al.  Direct Translation Model 2 , 2007, HLT-NAACL.

[5]  William J. Byrne,et al.  Hierarchical Phrase-Based Translation with Weighted Finite-State Transducers and Shallow-n Grammars , 2010, CL.

[6]  Joseph Olive,et al.  Handbook of Natural Language Processing and Machine Translation: DARPA Global Autonomous Language Exploitation , 2011 .

[7]  Heng Ji,et al.  Unsupervised Language-Independent Name Translation Mining from Wikipedia Infoboxes , 2011, ULNLP@EMNLP.

[8]  Jennifer Doyon,et al.  Automated Machine Translation Improvement Through Post-Editing Techniques: Analyst and Translator Experiments , 2008, AMTA.

[9]  Nizar Habash,et al.  Can Automatic Post-Editing Make MT More Meaningful , 2012, EAMT.

[10]  Philipp Koehn,et al.  Statistical Post-Editing on SYSTRAN‘s Rule-Based Translation System , 2007, WMT@ACL.

[11]  Alon Lavie,et al.  Meteor 1.3: Automatic Metric for Reliable Optimization and Evaluation of Machine Translation Systems , 2011, WMT@EMNLP.

[12]  Christopher D. Manning,et al.  Incorporating Non-local Information into Information Extraction Systems by Gibbs Sampling , 2005, ACL.

[13]  Nitin Madnani,et al.  Fluency, Adequacy, or HTER? Exploring Different Human Judgments with a Tunable MT Metric , 2009, WMT@EACL.

[14]  Cyril Goutte,et al.  Domain adaptation of MT systems through automatic post-editing , 2007, MTSUMMIT.

[15]  Michel Simard,et al.  Statistical Phrase-Based Post-Editing , 2007, NAACL.

[16]  Wen Wang,et al.  Using syntax in large-scale audio document translation , 2009, INTERSPEECH.

[17]  Roland Kuhn,et al.  Tighter Integration of Rule-Based and Statistical MT in Serial System Combination , 2008, COLING.

[18]  Ian H. Witten,et al.  The WEKA data mining software: an update , 2009, SKDD.

[19]  Dan Klein,et al.  Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network , 2003, NAACL.

[20]  Mark Dras,et al.  Choosing the Right Translation: A Syntactically Informed Classification Approach , 2008, COLING.

[21]  Kevin Knight,et al.  Automated Postediting of Documents , 1994, AAAI.

[22]  Lucia Specia,et al.  Predicting Machine Translation Adequacy , 2011, MTSUMMIT.

[23]  Marcello Federico,et al.  Match without a Referee: Evaluating MT Adequacy without Reference Translations , 2012, WMT@NAACL-HLT.

[24]  John Makhoul,et al.  Serial System Combination for Intergrating Rule-Based and Statistical Machine Translation , 2010 .

[25]  Wei-Yun Ma,et al.  Where’s the Verb? Correcting Machine Translation During Question Answering , 2009, ACL.