IRIS: English-Irish machine translation system

In this paper, we present our work on a case study between statistical machine translation (SMT) and rule-based machine translation (RBMT) systems on English-Indian language and Indian to Indian language perspective. Main objective of our study is to make a five-way performance comparison: such as, (a) SMT and RBMT; (b) SMT on English–Indian language; (c) RBMT on English–Indian language; (d) SMT on Indian to Indian language perspective; (e) RBMT on Indian to Indian language perspective. Through a detailed analysis, we describe the rule-based and the statistical machine translation system developments and its evaluations. Further, with a detailed error analysis, we point out the relative strengths and weaknesses of both the systems. The observations based on our study are: (a) SMT systems outperform RBMT; (b) In the case of SMT: English to Indian language MT systems perform better than Indian to English language MT systems; (c) In the case of RBMT: English to Indian language MT systems perform better than Indian to English language MT systems; (d) SMT systems perform better for Indian to Indian language MT systems compared to RBMT. Effectively, we shall see that even with a small amount of training corpus SMT system has many advantages for high-quality domain-specific machine translation over that of a rule-based counterpart.

Corpus-based machine translation (MT) has been the main approach to developing and implementing MT systems in both academia and the industry over the last three decades. In this field, the type and size of the corpus used for training MT engines have presented problems for both statistical MT (SMT) systems as well as neural MT (NMT) systems, being the two dominant corpusbased approaches. Moreover, language pairs such as Turkish-English have been understudied within this framework. This article aims to evaluate the translation quality in Turkish-to-English custom MT systems that have been trained on different corpus sizes and types. Two NMT engines and two SMT engines were trained on the KantanMT platform using two different training corpus types with either only domain-specific cardiology corpus or this corpus plus a mixed-domain corpus. The study conducted both automatic evaluations with metrics including BLEU, F-Measure and TER, as well as a comprehensive human evaluation with metrics including fluency, A/B test, and adequacy. Lastly, the study realized a separate, subjective terminology evaluation in order to investigate how differently MT systems handle terminology, as this is a crucial aspect for specific-domain text types such as cardiology. While the automatic evaluation results suggest the SMT engines to perform better than NMT engines, all human evaluators rated the mixed-domain NMT engine as the highest performing one. However, the terminology evaluation task demonstrated SMT to still be able to perform better and to commit less terminology errors, despite the industry and academia shifting toward NMT engines.

We report insights from translating Spanish conversational telephone speech into English text by cascading an automatic speech recognition (ASR) system with a statistical machine translation (SMT) system. The key new insight is that the informal register of conversational speech is a greater challenge for ASR than for SMT: the BLEU score for translating the reference transcript is 64%, but drops to 32% for translating automatic transcripts, whose word error rate (WER) is 40%. Several strategies are examined to mitigate the impact of ASR errors on the SMT output: (i) providing the ASR lattice, instead of the 1-best output, as input to the SMT system, (ii) training the SMT system on Spanish ASR output paired with English text, instead of Spanish reference transcripts, and (iii) improving the core ASR system. Each leads to consistent and complementary improvements in the SMT output. Compared to translating the 1-best output of an ASR system with 40% WER using an SMT system trained on Spanish reference transcripts, translating the output lattice of a better ASR system with 35% WER using an SMT system trained on ASR output improves BLEU from 32% to 38%.

In terms of language technology, Maithili is a resource-poor language. It is spoken in India and Nepal and is one of the 22 scheduled languages in India. Maithili has almost no language technology resource. English in India happens to be a dominant language in terms of content and usage. However, since more than 90% of Indians do not use English, a translation from English to Maithili becomes critical. An absence of basic tools in this language has affected resource creation of machine translation (MT). The present work discusses efforts for language technology resource (LTR) creation and divergence study for a statistical English-Maithili MT (EMMT) system. Creating any statistical MT (SMT) system requires sizeable parallel, aligned corpora for training and testing. Creating general-purpose source corpora for English language and creating translation equivalents with possible alignments require time and effort. The paper focuses on the data collection methods, cleaning, the size and structure of the text corpora, alignment and parallelization strategies, training, testing, and a study of divergence between the language pair.

A simplification–translation–restoration framework for domain adaptation in statistical machine translation: A case study in medical record translation

With the active participation of users in product review platforms, online consumer-generated content, and, more specifically, user-generated reviews, have become a clear reference in purchasing decision-making processes, which sometimes exceed the impact of advertising campaigns. A common feature of most tourism review platforms is the use of machine translation (MT) systems to immediately make reviews available to users in various languages. However, the quality of the MT output of these reviews varies greatly, primarily due to the subjective and unstructured nature of this digital genre. Different studies confirm that there are no universal quality rating scales. The assessment of MT output quality usually depends on factors such as the purpose of the text or the value given to the immediacy of the translation. New neural MT systems have been a revolution in the quality increase of the translated output; however, new lines of research are opening up to verify whether the quality of this new paradigm of MT can be assessed with the existing scales, mainly from previous rule-based systems and statistical translation, or whether it is necessary to develop new quality metrics specifically for these new intelligent systems. On the other hand, one of the questions that remain to be resolved in this new context of neural MT is whether the use of large amounts of textual data in the training of these systems is as effective as the use of less data but of higher quality and better-adjusted to the specialty and type of text for which it is used. Based on the hypothesis that each genre requires specific quality rating scales, this work identifies the error patterns and textual characteristics of online user reviews from a corpus-based approach analysis that will contribute to adapting quality rating scales to this specific digital genre.

Different Types of Machine Translation Systems in the Process of Translating Metaphors The two main topics discussed in this article are machine translation systems and metaphorical constructions translated by such systems. I briefly present the history of machine translation, its evolution and types of the most widespread engines, taking into account their architecture. I describe rule-based, statistical and neural machine translation, taking Translatica PWN, Google Statistical and Google Neural Machine Translation systems respectively as examples. Having discussed the notion of metaphor, I finally analyse examples of metaphorical expressions and idioms translated by the described engines, compare the final outcome and present my conclusions.

In a hospital, a medical summary is indispensable for both a clinician and a patient. However, it is written in English in some non-English native countries and becomes a barrier for a patient to read. In this paper we propose a framework for rapid acquisition of bilingual medical summaries using machine translation (MT) techniques. We describe a medical summary corpus and some terminological databases prepared for the framework. We then touch on the challenging issues of MT adapted from generic to specific domains, and propose a pattern translation scheme to achieve domain adaptation based on a background statistical MT system. We identify the significant patterns to capture the specific writing styles in a medical summary. The patterns are then translated with the involvements of doctors. Our major concern is to reduce the cost of translation and better allocate the efforts made by the domain experts. The experimental results show the proposed methods are effective in terms of the significance and diversity of the patterns. The approaches to integrate the mined patterns into background MT are also discussed.

Neither assigning similar priority to all phrases nor pruning out the incorrect phrases from the phrase table can improve the accuracy of machine translation. In this paper, we present a novel method for weight re-adjustment of phrase table in a statistical machine translation system. It learns the correct and incorrect phrases from bilingual corpora. Based on the syntactic phrase-level information, phrase table is updated with the weights estimated using probability distribution. Evaluation on English–Hindi technical domain corpora shows that our proposed method is more effective in producing better output in terms of BLEU, RIBES and NIST metrics. We shows that the proposed methods works well for other language pairs like Hindi–Konkani and Bengali–Hindi. Finally, we realised that this minor probabilistic change can improve the accuracy of the machine translation system a lot.

Though early successes of Statistical Machine Translation (SMT) systems are attributed in part to the explicit modelling of the interaction between any two source and target units, e.g., alignment, the recent Neural Machine Translation (NMT) systems resort to the attention which partially encodes the interaction for efficiency. In this paper, we employ Joint Representation that fully accounts for each possible interaction. We sidestep the inefficiency issue by refining representations with the proposed efficient attention operation. The resulting Reformer models offer a new Sequence-to-Sequence modelling paradigm besides the Encoder-Decoder framework and outperform the Transformer baseline in either the small scale IWSLT14 German-English, English-German and IWSLT15 Vietnamese-English or the large scale NIST12 Chinese-English translation tasks by about 1 BLEU point. We also propose a systematic model scaling approach, allowing the Reformer model to beat the state-of-the-art Transformer in IWSLT14 German-English and NIST12 Chinese-English with about 50% fewer parameters. The code is publicly available at https://github.com/lyy1994/reformer.

This chapter proposes algorithms for semisupervised learning. It translates sentences from the source language and then uses them to retrain the statistical machine translation (SMT) system in the hopes of getting a better translation system. It presents detailed experimental evaluations using French-English and Chinese-English data. The French-English translation task used bilingual data from the Europarl corpus, and monolingual data from the same domain as the test set which is drawn from the Canadian Hansard corpus. The Chinese-English task used bilingual data from the NIST large-data track and monolingual data from the Chinese Gigaword corpus.

Automatic translation of natural language is still (as of 2017) a long-standing but unmet promise. While advancing at a fast rate, the underlying methods are still far from actually being able to reliably capture syntax or semantics of arbitrary utterances of natural language, way off transporting the encoded meaning into a second language. However, it is possible to build useful translating machines when the target domain is well known and the machine is able to learn and adapt efficiently and promptly from new inputs. This is possible thanks to efficient and effective machine learning methods which can be applied to automatic translation. In this work we present and evaluate methods for three distinct scenarios: a) We develop algorithms that can learn from very large amounts of data by exploiting pairwise preferences defined over competing translations, which can be used to make a machine translation system robust to arbitrary texts from varied sources, but also enable it to learn effectively to adapt to new domains of data; b) We describe a method that is able to efficiently learn external models which adhere to fine-grained preferences that are extracted from a constricted selection of translated material, e.g. for adapting to users or groups of users in a computer-aided translation scenario; c) We develop methods for two machine translation paradigms, neural- and traditional statistical machine translation, to directly adapt to user-defined preferences in an interactive post-editing scenario, learning precisely adapted machine translation systems. In all of these settings, we show that machine translation can be made significantly more useful by careful optimization via preference learning.

This paper presents the system description of Machine Translation (MT) systems for English-Bangla language pair. Our goal was to create two benchmark MT systems that produce a better quality translation and comparatively higher evaluation score than existing MT systems for English to Bangla. In our experiments, we implemented two baseline MT systems using both statistical and neural methods for the said language pair. Our phrase-based statistical model and 2-layer LSTM neural model were trained and evaluated with a large dataset that is carefully pre-processed and contains unique training data to avoid biases from the cross-validation and test data. We achieved the highest scoring BLEU for our experiments with these setups. Furthermore, we improved the performance of the neural model using pre-trained embedding and synthetic monolingual data which are cutting-edge technology for neural models.

The paper presents a software re-engineering approach that is being used by a consortium of eleven academic institutions facilitated by a software engineering group to re-engineer previously developed natural language processing modules in laboratory environment, for developing eighteen machine translation (MT) systems covering nine pairs of Indic languages. This approach gives field deployable and maintainable MT systems (and modules) that can be reused as customizable components. The paper discusses its symbiotic approach of re-engineering modules, and also the experiences of developing MT systems in consortium mode. Following this approach, already seven MT systems have been developed and deployed on the web.KeywordsMachine TranslationLanguage ResourceLanguage PairSoftware Requirement SpecificationField DeployableThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Performance of statistical machine translation (SMT) systems relies on the availability of a large parallel corpus which is used to estimate translation probabilities. However, the generation of such corpus is a long and expensive process. In this paper, we introduce two methods for efficient selection of training data to be translated by humans. Our methods are motivated by active learning and aim to choose new data that adds maximal information to the currently available data pool. The first method uses a measure of disagreement between multiple SMT systems, whereas the second uses a perplexity criterion. We performed experiments on Chinese-English data in multiple domains and test sets. Our results show that we can select only one-fifth of the additional training data and achieve similar or better translation performance, compared to that of using all available data.