Báo cáo khoa học: "Statistical Machine Translation with Word- and Sentence-Aligned Parallel Corpora" potx

... sentence- and word-aligned train-ing material.5.4 Ratio of word- to sentence-aligned dataWe also varied the ratio of word-aligned to sentence-aligned data, and evaluated the AER and Bleu scores, and ... Statistical Machine Translation with Word- and Sentence-Aligned Parallel CorporaChris Callison-Burch David Talbot Miles OsborneSchool ... RateLambda20% word-aligned50% word-aligned70% word-aligned100% word-alignedFigure 2: The effect on AER of varying λ for a train-ing corpus of 16K sentence pairs with various pro-portions of word-alignmentsThus...

Tài liệu Báo cáo khoa học: "Statistical Machine Translation by Parsing" pptx

... tokenized set of -tuples of parallel sentences,it is always possible to estimate a word-to-word translation model (e.g., Och & Ney,2003).6A word-to-word translation model and a lexical-ized ... gram-mars and in their logics. They are all compatible with the same parsing semirings and search strate-gies. Therefore, we shall describe these algorithmsin terms of their underlying logics and ... synchronousstructures hidden in parallel texts. It turns out thatthese generalized parsers can do most of the workrequired to train and apply a syntax-aware statisti-cal machine translation system.1...

Báo cáo khoa học: "Statistical Machine Translation for Query Expansion in Answer Retrieval" pptx

... live) (with, with) (cat, cat) (allergies, food)(how, how) (to, to) (live, live) (with, with) (cat, cats) (allergies, allergies)paraphrasing (how, how) (to live, to live) (with cat, with cat) ... live, to live) (with cat, with cat) (allergies, allergies)(how, how) (to live with, to live with) (cat, feline) (allergies, allergies)(how to, how to) (live, living) (with cat, with cat) (allergies, ... to) (live, life) (with cat, with cat) (allergies, allergies)(how, way) (to live, to live) (with cat, with cat) (allergies, allergies)(how, how) (to live, to live) (with cat, with cat) (allergies,...

Báo cáo khoa học: "Statistical Machine Translation through Global Lexical Selection and Sentence Reconstruction" doc

... ambiguity in speech and other NLPtasks, outperforming generative models. Discrimi-native training has been used mainly for translation model combination (Och and Ney, 2002) and with the exception ... statis-tical machine translation, as exempliﬁed in (Brownet al., 1993), employs word-alignment algorithm(such as GIZA++ (Och and Ney, 2003)) that pro-vides local associations between source and targetwords. ... Bg.2.3 Bilingual Phrases and Local ReorderingWhile word-to-word translation only approximatesthe lexical selection process, phrase-to-phrase map-ping can greatly improve the translation of colloca-tions,...

Tài liệu Báo cáo khoa học: "Robust Machine Translation Evaluation with Entailment Features∗" pptx

... thatare semantically close but not identical. Banerjee and Lavie (2005) and Chan and Ng (2008) useWordNet, and Zhou et al. (2006) and Kauchak and Barzilay (2006) exploit large collections ofautomatically-extracted ... trans-lation and one or more reference translations. Theresulting scores are cheap and objective.However, studies such as Callison-Burch et al.(2006) have identiﬁed a number of problems with BLEU and ... Annual Meeting of the ACL and the 4th IJCNLP of the AFNLP, pages 297–305,Suntec, Singapore, 2-7 August 2009.c2009 ACL and AFNLPRobust Machine Translation Evaluation with Entailment Features∗Sebastian...

Tài liệu Báo cáo khoa học: "Collaborative Machine Translation Service for Scientiﬁc texts" pdf

... tokenisation) and the num-ber of words in the vocabulary (M and k stand for mil-lions and thousands, respectively).yielded a gain of more than 7 BLEU points, inboth domains (computer science and physics). ... article and can be re-trained to ﬁt tem-plates used for a speciﬁc conference or to extractadditional ﬁelds.4 Translation of Scientiﬁc TextsThe translation system used is a Hybrid Machine Translation ... compounds, pseudo-code, and enriched with domain bilingual terminology (seeSection 2). Large amounts of monolingual and parallel data are available to train a SMT systembetween French and English, but...

Báo cáo khoa học: "Is Machine Translation Ripe for Cross-lingual Sentiment Classiﬁcation" pdf

... 429–433,Portland, Oregon, June 19-24, 2011.c2011 Association for Computational LinguisticsIs Machine Translation Ripe for Cross-lingual Sentiment Classiﬁcation?Kevin Duh and Akinori Fujino and Masaaki ... advances in Machine Translation (MT)have brought forth a new paradigm for build-ing NLP applications in low-resource scenar-ios. To build a sentiment classiﬁer for alanguage with no labeled ... where labeled translations and testdata have some mismatch. Various prior workhave achieved positive results using this ap-proach.In this opinion piece, we take a step back and make some...

Báo cáo khoa học: "Improving Machine Translation of Null Subjects in Italian and Spanish" pot

... corpus and tuned on2,000 additional sentence pairs, and includes a 3-gram language model.Tables 3 and 4 show percentages of correct, in-correct and missing translations of personal and impersonal ... parsing and translation. Finally, we would like to use a larger corpus totrain the SPE component and compare the effectsof utilizing machine translations on the target sideversus human reference translations. ... GenGenP,2(1):64–78.Yu Chen and Andreas Eisele. 2010. HierarchicalHybrid Translation between English and German.In Proceedings of the 14th Annual Conference ofthe European Association for Machine Translation, pages...

Báo cáo khoa học: "Statistical Sense Disambiguation with Relatively Small Corpora Using Dictionary Definitions " ppt

... words that forms a statement, command, exclamation, or question, usu. contains a subject and a verb, and (in writing) begins with a capital letter and ends with one of the marks. ! ? conceptual ... representation and are readily available. 2 For instance, the LDOCE definitions of both offence and felony contain the word crime, and all of the definitions of sentence, fine and penalty contain ... co-occur with some words more often than with others. Similarly, looking at the meaning of the words, one should find that some concepts co-occur more often with some concepts than with others....

Báo cáo khoa học: "Learning to Translate with Source and Target Syntax" pot

... 2008.Syntax-driven learning of sub-sentential translation equivalents and translation rules from parsed parallel corpora. In Proc. SSST-2, pages 87–95.Yang Liu, Qun Liu, and Shouxun Lin. 2006. Tree-to-string ... template for statistical machine translation. In Proc. COLING-ACL 2006, pages609–616.Yang Liu, Yajuan L¨u, and Qun Liu. 2009. Improv-ing tree-to-tree translation with packed forests. InProc. ... Tan, and Sheng Li. 2008. A tree sequencealignment-based tree-to-tree translation model. InProc. ACL-08: HLT, pages 559–567.Andreas Zollmann and Ashish Venugopal. 2006. Syn-tax augmented machine...

Xem thêm