286 research outputs found
Free/Open-Source Machine Translation for the Low-Resource Languages of Spain (Invited Talk)
No full textWhile machine translation has historically been rule-based, that is, based on dictionaries and rules written by experts, most present-day machine translation is corpus-based. In the last few years, statistical machine translation, the dominant corpus-based approach, has been displaced by neural machine translation in most applications, in view of the better results reported, particularly for languages with very different syntax. But both statistical and neural machine translation need to be trained on large amounts of parallel data, that is, sentences in one language carefully paired with their translations in their other language, and this is a resource that may not be available for some low-resource languages. While some of the languages of Spain may be considered to be reasonably endowed with parallel corpora connecting them to Spanish or even to English - Basque, Catalan, Galician -, and are well-served with machine translation systems, there are many other languages which cannot afford them such as Aranese Occitan, Aragonese, or Asturian/Leonese. Fortunately, languages in this last group belong to the Romance language family, as Spanish does, and this makes translation from and into Spanish under a rule-based paradigm the only feasible approach. After describing briefly the main machine translation paradigms, I will describe the Apertium free/open-source rule-based machine translation platform, which has been used to build machine translation systems for these low-resource languages of Spain, indeed, sometimes the only ones available. The free/open-source setting has made linguistic data for these languages available for anyone in their linguistic communities to build other linguistic technologies for these low-resourced languages. For example, the Apertium family of bilingual and monolingual data has been converted into RDF and they have been made accessible on the Web as linked data
Editors’ foreword to the invited issue on SMT and NMT
No full textUntil quite recently, phrase-based statistical machine translation (PB-SMT) (Koehn et al. 2003, 2007; Koehn 2010) was indisputably the dominant paradigm in the field of MT. Papers suggesting how neural networks could be used for MT had been published twenty years ago (Chalmers 1990; Chrisman 1991; Castano and Casacuberta 1997; Forcada and Ñeco 1997; Ñeco and Forcada 1997), but the hardware around at the time was insufficient to support the amount of computation required for realistic experimentation
Experiments on domain adaptation for patent machine translation in the PLuTO project
No full textThe PLUTO1 project (Patent Language Translations Online) aims to provide a rapid solution for the online retrieval and translation of patent documents through the integration of a number of existing state-of-the-art components provided by the project partners. The paper presents some of the experiments on patent domain adaptation of the Machine Translation (MT) systems used in the PLuTO project. The experiments use the International Patent Classification for domain adaptation and are focused on the English–French language pair
Recurrent neural networks can learn simple, approximate regular languages
No full textA number of researchers have shown that discrete-time recurrent neural networks (DTRNN) are capable of inferring deterministic finite automata from sets of example and counterexample strings; however, discrete algorithmic methods are much better at this task and clearly outperform DTRNN in terms of space and time complexity. We show how DTRNN may be used to learn not the exact language that explains the whole learning set but an approximate and much simpler language that explains a great majority of the examples by using simpler rules. This is accomplished by gradually varying the error function in such a way that the DTRNN is eventually allowed to classify clearly but incorrectly those strings that it has found to be difficult to learn, which are treated as exceptions. The results show that in this way, the DTRNN usually manages to learn a simplified approximate language
Meta-Evaluation of a Diagnostic Quality Metric for Machine Translation
No full textDiagnostic evaluation of machine translation
(MT) is an approach to evaluation that
provides finer-grained information compared
to state-of-the-art automatic metrics.
This paper evaluates DELiC4MT, a diagnostic
metric that assesses the performance
of MT systems on user-defined linguistic
phenomena. We present the results obtained
using this diagnostic metric when
evaluating three MT systems that translate
from English to French, with a comparison
against both human judgements and
a set of representative automatic evaluation
metrics. In addition, as the diagnostic
metric relies on word alignments, the
paper compares the margin of error in diagnostic
evaluation when using automatic
word alignments as opposed to gold standard
manual alignments. We observed that
this diagnostic metric is capable of accurately
reflecting translation quality, can be
used reliably with automatic word alignments
and, in general, correlates well with
automatic metrics and, more importantly,
with human judgements
Hybrid rule-based - example-based MT: feeding apertium with sub-sentential translation units
No full textThis paper describes a hybrid machine translation (MT) approach that consists of integrating bilingual chunks (sub-sentential translation units) obtained from parallel corpora into an MT system built using the Apertium free/open-source rule-based machine translation platform, which uses a shallow-transfer translation approach. In the integration of bilingual chunks, special care has been
taken so as not to break the application of the existing Apertium structural transfer rules, since this would increase the number of ungrammatical translations. The method consists of (i) the application of a dynamic-programming algorithm to compute the best translation coverage of the input sentence given the collection of bilingual chunks available; (ii) the translation of the input sentence as usual by Apertium; and (iii) the application of a language model to choose one of the possible translations for each of the bilingual chunks detected. Results are reported for the translation from English-to-Spanish, and vice versa, when marker-based bilingual chunks automatically obtained from parallel
corpora are used
Towards Optimizing MT for Post-Editing Effort: Can BLEU Still Be Useful?
No full textWe propose a simple, linear-combination automatic evaluation measure (AEM) to approximate post-editing (PE) effort. Effort is measured both as PE time and as the number of PE operations performed. The ultimate goal is to define an AEM that can be used to optimize machine translation (MT) systems to minimize PE effort, but without having to perform unfeasible repeated PE during optimization. As PE effort is expected to be an extensive magnitude (i.e., one growing linearly with the sentence length and which may be simply added to represent the effort for a set of sentences), we use a linear combination of extensive and pseudo-extensive features. One such pseudo-extensive feature, 1–BLEU times the length of the reference, proves to be almost as good a predictor of PE effort as the best combination of extensive features. Surprisingly, effort predictors computed using independently obtained reference translations perform reasonably close to those using actual post-edited references. In the early stage of this research and given the inherent complexity of carrying out experiments with professional post-editors, we decided to carry out an automatic evaluation of the AEMs proposed rather than a manual evaluation to measure the effort needed to post-edit the output of an MT system tuned on these AEMs. The results obtained seem to support current tuning practice using BLEU, yet pointing at some limitations. Apart from this intrinsic evaluation, an extrinsic evaluation was also carried out in which the AEMs proposed were used to build synthetic training corpora for MT quality estimation, with results comparable to those obtained when training with measured PE efforts.Work supported by the Spanish government through project EFFORTUNE (TIN2015-69632-R) and through grant PRX16/00043 for Mikel L. Forcada, and by the European Commission through QT21 project (H2020 No. 645452)
Making sense of neural machine translation
No full textThe last few years have witnessed a surge in the interest of a new machine translation paradigm: neural machine translation (NMT). Neural machine translation is starting to displace its corpus-based predecessor, statistical machine translation (SMT). In this paper, I introduce NMT, and explain in detail, without the mathematical complexity, how neural machine translation systems work, how they are trained, and their main differences with SMT systems. The paper will try to decipher NMT jargon such as “distributed representations”, “deep learning”, “word embeddings”, “vectors”, “layers”, “weights”, “encoder”, “decoder”, and “attention”, and build upon these concepts, so that individual translators and professionals working for the translation industry as well as students and academics in translation studies can make sense of this new technology and know what to expect from it. Aspects such as how NMT output differs from SMT, and the hardware and software requirements of NMT, both at training time and at run time, on the translation industry, will be discussed.This work was performed while the author was on sabbatical leave at the University of Sheffield and the University of Edinburgh: the author thanks Universitat d’Alacant and the Spanish Ministry of Education, Culture and Sport (grant number PRX16/00043) for support during this leave
Ontolex-lemon and TIAD versions of Apertium English-Kazakh dictionary
No full textOntoLex-lemon and TSV conversion of Apertium Bidix. For more details, see https://www.aclweb.org/anthology/2020.lrec-1.401/
Authors of the original data:
(c) 2012 Assem Shormakova
(c) 2012 Mikel L. Forcada
(c) 2012 Ilnar Salimzyanov
(c) 2011-2012 Jonathan N. Washington
(c) 2011-2012 Francis M. Tyers
(c) 2011 Nathan Maxso
Using unsupervised corpus-based methods to build rule-based machine translation systems
No full textTesis doctoral en Informática realizada en la Universitat d’Alacant por Felipe Sánchez Martínez bajo la dirección de los doctores Juan Antonio Pérez Ortiz y Mikel L. Forcada. La defensa de la tesis tuvo lugar el 30 de junio de 2008 ante el tribunal formado por los doctores Rafael C. Carrasco (Univ. d’Alacant), Lluís Padró y Lluís Màrquez (Univ. Politècnica de Catalunya), Harold Somers (Univ. of Manchester) y Andy Way (Dublin City Univ.). La calificación obtenida fue Sobresaliente Cum Laude por unanimidad, con mención de Doctor Europeo.PhD thesis in Computer Engineering written by Felipe Sánchez-Martínez at Universitat d’Alacant under the joint supervision of Dr. Juan Antonio Pérez-Ortiz and Dr. Mikel L. Forcada. Author was examined on June 30th , 2008 by the committee formed by Dr. Rafael C. Carrasco (Univ. d’Alacant), Dr. Lluís Padró and Dr. Lluís Màrquez (Univ. Politècnica de Catalunya), Dr. Harold Somers (Univ. of Manchester) and Dr. Andy Way (Dublin City Univ.). The grade obtained was Sobresaliente Cum Laude (highest mark), with the European Doctor mention.Tesis financiada por el Ministerio de Educación y Ciencia y el Fondo Social Europeo a través de la ayuda a la investigación BES-2004-4711
- …
