Víctor M. Sánchez-Cartagena

Tradubi: Open-Source Social Translation for the Apertium Machine Translation Platform

Tradubi: Open-Source Social Translation for the Apertium Machine Translation Platform Massive online collaboration could become a winning strategy to tear down the language barriers on the web, and in order for this to happen appropriate computer tools, like reliable machine translation systems and friendly postediting interfaces, should be widely available. However, community collaboration should not only involve the postediting of machine translations, but also the creation of the linguistic resources needed to improve the translation engines. In this paper we introduce Tradubi, a free/open-source web application for social translation, whose aim is, firstly, to build a platform for collaboratively customising and improving rule-based machine translation systems and, secondly, to offer an environment for the postediting and subsequent sharing of raw machine translations. Currently, Tradubi is built upon the free/open-source Apertium machine translation engine. The application can be accessed at tradubi.com or downloaded and installed on a different server.

A generalised alignment template formalism and its application to the inference of shallow-transfer machine translation rules from scarce bilingual corpora

HighlightsNew approach to infer shallow-transfer rules from scarce parallel corpora.New rule formalism permits strong generalisation over the parallel corpus.First approach in which rule learning is rewritten as a global minimisation problem.Translation quality improves over previous approach with a smaller number of rules.Translation quality outperforms hand-coded rules for some language pairs. Statistical and rule-based methods are complementary approaches to machine translation (MT) that have different strengths and weaknesses. This complementarity has, over the last few years, resulted in the consolidation of a growing interest in hybrid systems that combine both data-driven and linguistic approaches. In this paper, we address the situation in which the amount of bilingual resources that is available for a particular language pair is not sufficiently large to train a competitive statistical MT system, but the cost and slow development cycles of rule-based MT systems cannot be afforded either. In this context, we formalise a new method that uses scarce parallel corpora to automatically infer a set of shallow-transfer rules to be integrated into a rule-based MT system, thus avoiding the need for human experts to handcraft these rules.Our work is based on the alignment template approach to phrase-based statistical MT, but the definition of the alignment template is extended to encompass different generalisation levels. It is also greatly inspired by the work of Sanchez-Martinez and Forcada (2009) in which alignment templates were also considered for shallow-transfer rule inference. However, our approach overcomes many relevant limitations of that work, principally those related to the inability to find the correct generalisation level for the alignment templates, and to select the subset of alignment templates that ensures an adequate segmentation of the input sentences by the rules eventually obtained. Unlike previous approaches in literature, our formalism does not require linguistic knowledge about the languages involved in the translation. Moreover, it is the first time that conflicts between rules are resolved by choosing the most appropriate ones according to a global minimisation function rather than proceeding in a pairwise greedy fashion.Experiments conducted using five different language pairs with the free/open-source rule-based MT platform Apertium show that translation quality significantly improves when compared to the method proposed by Sanchez-Martinez and Forcada (2009), and is close to that obtained using handcrafted rules. For some language pairs, our approach is even able to outperform them. Moreover, the resulting number of rules is considerably smaller, which eases human revision and maintenance.

ScaleMT: a Free/Open-Source Framework for Building Scalable Machine Translation Web Services

ScaleMT: a Free/Open-Source Framework for Building Scalable Machine Translation Web Services Machine translation web services usage is growing amazingly mainly because of the translation quality and reliability of the service provided by the Google Ajax Language API. To allow the open-source machine translation projects to compete with Googles one and gain visibility on the internet, we have developed ScaleMT: a free/open-source framework that exposes existing machine translation engines as public web services. This framework is highly scalable as it can run coordinately on many servers, efficiently managing the resources needed by the engines, and its API is compatible with Googles one. ScaleMT is based on previous efforts to build a web service for the Apertium machine translation toolbox, but we have also tested it with Matxin, another free/open-source transfer-based machine translation engine. Additionally, we have compared ScaleMT to an alternative web service implementation for Apertium, obtaining satisfactory results.

Fine-Grained Human Evaluation of Neural Versus Phrase-Based Machine Translation

Abstract We compare three approaches to statistical machine translation (pure phrase-based, factored phrase-based and neural) by performing a fine-grained manual evaluation via error annotation of the systems’ outputs. The error types in our annotation are compliant with the multidimensional quality metrics (MQM), and the annotation is performed by two annotators. Inter-annotator agreement is high for such a task, and results show that the best performing system (neural) reduces the errors produced by the worst system (phrase-based) by 54%.

Abu-MaTran at WMT 2016 Translation Task: Deep Learning, Morphological Segmentation and Tuning on Character Sequences

This paper presents the systems submitted by the Abu-MaTran project to the Englishto-Finnish language pair at the WMT 2016 news translation task. We applied morphological segmentation and deep learning in order to address (i) the data scarcity problem caused by the lack of in-domain parallel data in the constrained task and (ii) the complex morphology of Finnish. We submitted a neural machine translation system, a statistical machine translation system reranked with a neural language model and the combination of their outputs tuned on character sequences. The combination and the neural system were ranked first and second respectively according to automatic evaluation metrics and tied for the first place in the human evaluation.

Abu-MaTran at WMT 2014 Translation Task: Two-step Data Selection and RBMT-Style Synthetic Rules

This paper presents the machine translation systems submitted by the AbuMaTran project to the WMT 2014 translation task. The language pair concerned is English‐French with a focus on French as the target language. The French to English translation direction is also considered, based on the word alignment computed in the other direction. Large language and translation models are built using all the datasets provided by the shared task organisers, as well as the monolingual data from LDC. To build the translation models, we apply a two-step data selection method based on bilingual crossentropy difference and vocabulary saturation, considering each parallel corpus individually. Synthetic translation rules are extracted from the development sets and used to train another translation model. We then interpolate the translation models, minimising the perplexity on the development sets, to obtain our final SMT system. Our submission for the English to French translation task was ranked second amongst nine teams and a total of twenty submissions.

Integrating rules and dictionaries from shallow-transfer machine translation into phrase-based statistical machine translation

We describe a hybridisation strategy whose objective is to integrate linguistic resources from shallow-transfer rule-based machine translation (RBMT) into phrase-based statistical machine translation (PBSMT). It basically consists of enriching the phrase table of a PBSMT system with bilingual phrase pairs matching transfer rules and dictionary entries from a shallow-transfer RBMT system. This new strategy takes advantage of how the linguistic resources are used by the RBMT system to segment the source-language sentences to be translated, and overcomes the limitations of existing hybrid approaches that treat the RBMT systems as a black box. Experimental results confirm that our approach delivers translations of higher quality than existing ones, and that it is specially useful when the parallel corpus available for training the SMT system is small or when translating out-of-domain texts that are well covered by the RBMT dictionaries. A combination of this approach with a recently proposed unsupervised shallow-transfer rule inference algorithm results in a significantly greater translation quality than that of a baseline PBSMT; in this case, the only hand-crafted resource used are the dictionaries commonly used in RBMT. Moreover, the translation quality achieved by the hybrid system built with automatically inferred rules is similar to that obtained by those built with hand-crafted rules.

The UA-Prompsit hybrid machine translation system for the 2014 Workshop on Statistical Machine Translation

This paper describes the system jointly developed by members of the Departament de Llenguatges i Sistemes Informatics at Universitat d’Alacant and the Prompsit Language Engineering company for the shared translation task of the 2014 Workshop on Statistical Machine Translation. We present a phrase-based statistical machine translation system whose phrase table is enriched with information obtained from dictionaries and shallowtransfer rules like those used in rule-based machine translation. The novelty of our approach lies in the fact that the transfer rules used were not written by humans, but automatically inferred from a parallel cor-

RuLearn: an Open-source Toolkit for the Automatic Inference of Shallow-transfer Rules for Machine Translation

Abstract This paper presents ruLearn, an open-source toolkit for the automatic inference of rules for shallow-transfer machine translation from scarce parallel corpora and morphological dictionaries. ruLearn will make rule-based machine translation a very appealing alternative for under-resourced language pairs because it avoids the need for human experts to handcraft transfer rules and requires, in contrast to statistical machine translation, a small amount of parallel corpora (a few hundred parallel sentences proved to be sufficient). The inference algorithm implemented by ruLearn has been recently published by the same authors in Computer Speech & Language (volume 32). It is able to produce rules whose translation quality is similar to that obtained by using hand-crafted rules. ruLearn generates rules that are ready for their use in the Apertium platform, although they can be easily adapted to other platforms. When the rules produced by ruLearn are used together with a hybridisation strategy for integrating linguistic resources from shallow-transfer rule-based machine translation into phrase-based statistical machine translation (published by the same authors in Journal of Artificial Intelligence Research, volume 55), they help to mitigate data sparseness. This paper also shows how to use ruLearn and describes its implementation.

Quantitative fine-grained human evaluation of machine translation systems: a case study on English to Croatian

This paper presents a quantitative fine-grained manual evaluation approach to comparing the performance of different machine translation (MT) systems. We build upon the well-established multidimensional quality metrics (MQM) error taxonomy and implement a novel method that assesses whether the differences in performance for MQM error types between different MT systems are statistically significant. We conduct a case study for English-to-Croatian, a language direction that involves translating into a morphologically rich language, for which we compare three MT systems belonging to different paradigms: pure phrase-based, factored phrase-based and neural. First, we design an MQM-compliant error taxonomy tailored to the relevant linguistic phenomena of Slavic languages, which made the annotation process feasible and accurate. Errors in MT outputs were then annotated by two annotators following this taxonomy. Subsequently, we carried out a statistical analysis which showed that the best-performing system (neural) reduces the errors produced by the worst system (pure phrase-based) by more than half (54%). Moreover, we conducted an additional analysis of agreement errors in which we distinguished between short (phrase-level) and long distance (sentence-level) errors. We discovered that phrase-based MT approaches are of limited use for long distance agreement phenomena, for which neural MT was found to be especially effective.