Machine Translation Problem Research Articles

한-스페인어 기계번역에 나타나는 제반 특징에 관한 연구

This paper aims to analyze the general characteristics of Korean-Spanish machine translation and vice versa. In spite of recent developments in artificial intelligence, there still are numerous errors in lexical and syntactic structure in machine translation. One of the most important reasons for these errors in machine translation is word-for-word translation without considering the linguistic differences of Korean and Spanish. For the purpose of our analysis, we focused on two grammatical categories, noun phrases and verbal phrases in the 100 cases. In the analysis of noun phrases, we investigated the problems on plurality and definiteness in Koean and Spanish. In our analysis of verb phrases in context, we find that the mood and tense cause some problems in machine translation.. We believe that this paper can contribute to a comparative analysis of human and machine translation in the near future. We insist that we must find some feedback system to improve the quality of machine translation through systematic and coherent research cooperation.

Improving Machine Translation Quality with Denoising Autoencoder and Pre-Ordering

The problems in machine translation are related to the characteristics of a family of languages, especially syntactic divergences between languages. In the translation task, having both source and target languages in the same language family is a luxury that cannot be relied upon. The trained models for the task must overcome such differences either through manual augmentations or automatically inferred capacity built into the model design. In this work, we investigated the impact of multiple methods of differing word orders during translation and further experimented in assimilating the source languages syntax to the target word order using pre-ordering. We focused on the field of extremely low-resource scenarios. We also conducted experiments on practical data augmentation techniques that support the reordering capacity of the models through varying the target objectives, adding the secondary goal of removing noises or reordering broken input sequences. In particular, we propose methods to improve translat on quality with the denoising autoencoder in Neural Machine Translation (NMT) and pre-ordering method in Phrase-based Statistical Machine Translation (PBSMT). The experiments with a number of English-Vietnamese pairs show the improvement in BLEU scores as compared to both the NMT and SMT systems.

Retracted] Design of Intelligent Recognition English Translation Model Based on Deep Learning

Nowadays, the intercommunication and translation of global languages has become an indispensable condition for friendly communication among human beings around the world. The advancement of computer technology developed the machine translation from academic research to industrial applications. Additionally, a new and popular branch of machine learning is deep learning which has achieved excellent results in research fields such as natural language processing. This paper improved the performance of machine translation based on deep learning network and studied the intelligent recognition of English‐Chinese machine translation models. This research mainly focused on solving out‐of‐vocabulary (OOV) problem of machine translation on unregistered words and rare words. Moreover, it combined stemming technology and data compression algorithm Byte Pair Encoding (BPE) and proposed a different subword‐based word sequence segmentation method. Using this method, the English text is segmented into word sequences composed of subword units, and, at the same time, the Chinese text is segmented into character sequences composed of Chinese characters using unigram. Secondly, the current research also prevented the decoder from experiencing incomplete translation. Furthermore, it adopted a deep‐attention mechanism that can improve the decoder's ability to obtain context information. Inspired by the traditional attention calculation process, this work uses a two‐layer calculation structure in the improved attention to focus on the connection between the context vectors at different moments of the decoder. Based on the neural machine translation model Google Neural Machine Translation (GNMT), this paper conducted experimental analysis on the above improved methods on three different scale datasets. Experimental results verified that the improved method can solve OOV problem and improve accuracy of model translation.

Concept recognition as a machine translation problem

BackgroundAutomated assignment of specific ontology concepts to mentions in text is a critical task in biomedical natural language processing, and the subject of many open shared tasks. Although the current state of the art involves the use of neural network language models as a post-processing step, the very large number of ontology classes to be recognized and the limited amount of gold-standard training data has impeded the creation of end-to-end systems based entirely on machine learning. Recently, Hailu et al. recast the concept recognition problem as a type of machine translation and demonstrated that sequence-to-sequence machine learning models have the potential to outperform multi-class classification approaches.MethodsWe systematically characterize the factors that contribute to the accuracy and efficiency of several approaches to sequence-to-sequence machine learning through extensive studies of alternative methods and hyperparameter selections. We not only identify the best-performing systems and parameters across a wide variety of ontologies but also provide insights into the widely varying resource requirements and hyperparameter robustness of alternative approaches. Analysis of the strengths and weaknesses of such systems suggest promising avenues for future improvements as well as design choices that can increase computational efficiency with small costs in performance.ResultsBidirectional encoder representations from transformers for biomedical text mining (BioBERT) for span detection along with the open-source toolkit for neural machine translation (OpenNMT) for concept normalization achieve state-of-the-art performance for most ontologies annotated in the CRAFT Corpus. This approach uses substantially fewer computational resources, including hardware, memory, and time than several alternative approaches.ConclusionsMachine translation is a promising avenue for fully machine-learning-based concept recognition that achieves state-of-the-art results on the CRAFT Corpus, evaluated via a direct comparison to previous results from the 2019 CRAFT shared task. Experiments illuminating the reasons for the surprisingly good performance of sequence-to-sequence methods targeting ontology identifiers suggest that further progress may be possible by mapping to alternative target concept representations. All code and models can be found at: https://github.com/UCDenver-ccp/Concept-Recognition-as-Translation.

Parallel Algorithm of Hierarchical Phrase Machine Translation Based on Distributed Network Memory

Machine translation has developed rapidly. But there are some problems in machine translation, such as good reading, unable to reflect the mood and context, and even some language machines can not recognize. In order to improve the quality of translation, this paper uses the SSCI method to improve the quality of translation. It is found that the translation quality of hierarchical phrases is significantly improved after using the parallel algorithm of machine translation, which is about 9% higher than before, and the problem of context free grammar is also solved. The research also found that the use of parallel algorithm can effectively reduce the network memory occupation, the original 10 character content, after using the parallel algorithm, only need to occupy 8 characters, the optimization reaches 20%. This means that the parallel algorithm of hierarchical phrase machine translation based on distributed network memory can play a very important role in machine translation.

THE TRANSLATION QUALITY PROBLEMS OF MACHINE TRANSLATION SYSTEMS FOR THE KAZAKH LANGUAGE

Kazakh language is related to languages with complex morphology and syntax. Today, most machine translation systems consider Kazakh language too, like Google, Yandex, Prompt, etc. This article describes the errors, shortcomings, and problems of machine translation (MT) into the Kazakh language. To analyze errors in machine translation into the Kazakh language, the most popular electronic translation programs were selected. When translating from Russian and English into Kazakh (and vice versa), various errors may occur, since the Kazakh language is dierent from other languages, and has special characteristics. To compare the results, an empirical method was used, namely, monitoring and testing the translation results of machine translation systems. Considering the results of statistical methods, the rule-based and neural networks based methods in machine translations were also analyzed. The practical signicance of the study lies in the development of recommendations for the identication and elimination of errors when editing the results of MT. The scientic signicance of the study lies in the fact that for the rst time errors and inaccuracies arising from machine translation of the Kazakh language have been systematized. The assessment of the quality of MT is also presented. The research carried out in this article will be used for the post-editing problem in machine translation.

Improving Neural Machine Translation with AMR Semantic Graphs

The Seq2Seq model and its variants (ConvSeq2Seq and Transformer) emerge as a promising novel solution to the machine translation problem. However, these models only focus on exploiting knowledge from bilingual sentences without paying much attention to utilizing external linguistic knowledge sources such as semantic representations. Not only do semantic representations can help preserve meaning but they also minimize the data sparsity problem. However, to date, semantic information remains rarely integrated into machine translation models. In this study, we examine the effect of abstract meaning representation (AMR) semantic graphs in different machine translation models. Experimental results on the IWSLT15 English-Vietnamese dataset have proven the efficiency of the proposed model, expanding the use of external language knowledge sources to significantly improve the performance of machine translation models, especially in the application of low-resource language pairs.

Towards Fully Automated Manga Translation

We tackle the problem of machine translation of manga, Japanese comics. Manga translation involves two important problems in machine translation: context-aware and multimodal translation. Since text and images are mixed up in an unstructured fashion in Manga, obtaining context from the image is essential for manga translation. However, it is still an open problem how to extract context from image and integrate into MT models. In addition, corpus and benchmarks to train and evaluate such model is currently unavailable. In this paper, we make the following four contributions that establishes the foundation of manga translation research. First, we propose multimodal context-aware translation framework. We are the first to incorporate context information obtained from manga image. It enables us to translate texts in speech bubbles that cannot be translated without using context information (e.g., texts in other speech bubbles, gender of speakers, etc.). Second, for training the model, we propose the approach to automatic corpus construction from pairs of original manga and their translations, by which large parallel corpus can be constructed without any manual labeling. Third, we created a new benchmark to evaluate manga translation. Finally, on top of our proposed methods, we devised a first compleheisive system for fully automated manga translation.

Regression Loss in Transformer-based Supervised Neural Machine Translation

Transformer-based model has achieved human-level performance in supervised neural machine translation (SNMT), much better than the models based on recurrent neural networks (RNNs) or convolutional neural network (CNN). The original Transformer-based model is trained through maximum likelihood estimation (MLE), which regards the machine translation task as a multilabel classification problem and takes the sum of the cross entropy loss of all the target tokens as the loss function. However, this model assumes that token generation is partially independent, without realizing that tokens are the components of a sequence. To solve the problem, this paper proposes a semantic regression loss for Transformer training, treating the generated sequence as a global. Upon finding that the semantic difference is proportional to candidate-reference distance, the authors considered the machine translation problem as a multi-task problem, and took the linear combination of cross entropy loss and semantic regression loss as the overall loss function. The semantic regression loss was proved to significantly enhance SNMT performance, with a slight reduction in convergence speed.

Molecular optimization by capturing chemist\u2019s intuition using deep neural networks

A main challenge in drug discovery is finding molecules with a desirable balance of multiple properties. Here, we focus on the task of molecular optimization, where the goal is to optimize a given starting molecule towards desirable properties. This task can be framed as a machine translation problem in natural language processing, where in our case, a molecule is translated into a molecule with optimized properties based on the SMILES representation. Typically, chemists would use their intuition to suggest chemical transformations for the starting molecule being optimized. A widely used strategy is the concept of matched molecular pairs where two molecules differ by a single transformation. We seek to capture the chemist’s intuition from matched molecular pairs using machine translation models. Specifically, the sequence-to-sequence model with attention mechanism, and the Transformer model are employed to generate molecules with desirable properties. As a proof of concept, three ADMET properties are optimized simultaneously: logD, solubility, and clearance, which are important properties of a drug. Since desirable properties often vary from project to project, the user-specified desirable property changes are incorporated into the input as an additional condition together with the starting molecules being optimized. Thus, the models can be guided to generate molecules satisfying the desirable properties. Additionally, we compare the two machine translation models based on the SMILES representation, with a graph-to-graph translation model HierG2G, which has shown the state-of-the-art performance in molecular optimization. Our results show that the Transformer can generate more molecules with desirable properties by making small modifications to the given starting molecules, which can be intuitive to chemists. A further enrichment of diverse molecules can be achieved by using an ensemble of models.

Transformer neural network for protein-specific de novo drug generation as a machine translation problem

Drug discovery for a protein target is a very laborious, long and costly process. Machine learning approaches and, in particular, deep generative networks can substantially reduce development time and costs. However, the majority of methods imply prior knowledge of protein binders, their physicochemical characteristics or the three-dimensional structure of the protein. The method proposed in this work generates novel molecules with predicted ability to bind a target protein by relying on its amino acid sequence only. We consider target-specific de novo drug design as a translational problem between the amino acid “language” and simplified molecular input line entry system representation of the molecule. To tackle this problem, we apply Transformer neural network architecture, a state-of-the-art approach in sequence transduction tasks. Transformer is based on a self-attention technique, which allows the capture of long-range dependencies between items in sequence. The model generates realistic diverse compounds with structural novelty. The computed physicochemical properties and common metrics used in drug discovery fall within the plausible drug-like range of values.

Valid, Plausible, and Diverse Retrosynthesis Using Tied Two-Way Transformers with Latent Variables.

Retrosynthesis is an essential task in organic chemistry for identifying the synthesis pathways of newly discovered materials, and with the recent advances in deep learning, there have been growing attempts to solve the retrosynthesis problem through transformer models, which are the state-of-the-art in neural machine translation, by converting the problem into a machine translation problem. However, the pure transformer provides unsatisfactory results that lack grammatical validity, chemical plausibility, and diversity in reactant candidates. In this study, we develop tied two-way transformers with latent modeling to solve those problems using cycle consistency checks, parameter sharing, and multinomial latent variables. Experimental results obtained using public and in-house datasets demonstrate that the proposed model improves the retrosynthesis accuracy, grammatical error, and diversity, and qualitative evaluation results verify its ability to suggest valid and plausible results.

Importance of anaphora resolution in machine translation

In this era of continuously progressing Information Technology, information exchange is taking place in various domains at a faster pace. However, availability of huge information, diversity of knowledge sources and variations in structure of languages, leads to some hurdles in its effective utilization. Use of Machine Translation is a good enough solution at this stage, which can remove several constraints up-to some extent. Also, because of the easy and economic Internet accessibility, Machine Translation is becoming crucial. But several technical constraints are there in Machine Translation process that must be resolved for qualitative translation. Anaphora resolution is one of the major and complex problems of Machine Translation. Many researchers and research institute working in the field of language technology and translation are continuously striving to resolve the problem. Depending upon type of language there may be various classes of anaphor.Therefore, depending upon the type of source and target language, type of anaphor and anaphor resolution process also may differ. This paper is an attempt to go in detail and reviewing of currently available theoretical approaches and algorithms of anaphora resolution that are being used in various machine translation models.

Deep Learning-based Roman-Urdu to Urdu Transliteration

Attention-based encoder-decoder models have superseded conventional techniques due to their unmatched performance on many neural machine translation problems. Usually, the encoders and decoders are two recurrent neural networks where the decoder is directed to focus on relevant parts of the source language using attention mechanism. This data-driven approach leads to generic and scalable solutions with no reliance on manual hand-crafted features. To the best of our knowledge, none of the modern machine translation approaches has been applied to address the research problem of Urdu machine transliteration. Ours is the first attempt to apply the deep neural network-based encoder-decoder using attention mechanism to address the aforementioned problem using Roman-Urdu and Urdu parallel corpus. To this end, we present (i) the first ever Roman-Urdu to Urdu parallel corpus of 1.1 million sentences, (ii) three state of the art encoder-decoder models, and (iii) a detailed empirical analysis of these three models on the Roman-Urdu to Urdu parallel corpus. Overall, attention-based model gives state-of-the-art performance with the benchmark of 70 BLEU score. Our qualitative experimental evaluation shows that our models generate coherent transliterations which are grammatically and logically correct.

Challenges in Machine Translation For Teaching Arabic Language/ الترجمة الحاسوبية والتحديات في التعليم اللغة العربية

The study aims to search for syntax errors in machine translation. The study focuses on challenges in English Arabic Machine translation, types of errors in translation software, the most developed technologies in machine translation, and arbitrary Arabic text generating problems generally and in the Arabic language. This study use qualitative research design, qualitative data, drawn from our pragmatic analysis are presented, from perceptions of language learners and tutors in relation of Machine translation. The article discusses the disappointing fact that most intelligent machines are built to produce woodenly literal or paraphrased translations, but in Arabic, the errors are in high numbers. The paper results show an urgent need to build the machine language infrastructure in the Arabic language field to help people achieve accurate translation. To overcome the problems of machine translation, we recommend that we enter enough vocabulary and linguistic texts in the machine.

Facile Solutions to the Problems Associated with Chemical Information and Mathematical Symbolism While Using Machine Translation Tools.

Advances in computer-aided translation technology have made tremendous progress in accuracy in the past few years. Chemical Abstracts Service of the American Chemical Society summarizes scientific works from more than 50 languages and allows the users to search papers in nine selected languages. Currently, only the abstracts are rendered into English by human experts or by machine translation because full text translation of millions of articles is beyond the human capacity today. An English translation of a research paper, scientific book, or patent is often required for research, data mining, and for historical purposes from various foreign languages. Many fundamental papers in chemistry, quantum chemistry, physics, and mathematics contain a significant number of chemical or mathematical equations. One of the major known problems in machine translation of such symbolically dense texts is incorrect or meaningless output. This article describes how to optimize the existing machine translation tools to read foreign language papers embedded with chemical/mathematical equations. German and French languages have been selected for illustrative purposes for English translation. Direct upload of text with extensive symbolism is possible with certain services, but this also occasionally produces erroneous rendition into English. A facile solution to the associated problems with embedded equations and mathematical formulas is replacing the equations and notations with "dummy" variables. The placeholder or dummy symbols can be removed after translation, and the original equations are substituted again. This approach, which can be automated in future, relies on the idea that chemical formulas and mathematical notations are universal. Following the guidelines in the article, excellent translations can be produced from a text having interspersed equations and chemical symbols.

Machine Phraseological Translation of Scientific-Technical Texts Based on the Model of Generalized Syntagmas

This article proposes a new solution to the problem of phraseological machine translation of texts based on the technology of “semantic translation constructor.” This technology is based on the concept of generalized syntagmas, which is based on the proposition that the syntactic structure of a sentence or its fragments (syntactic constructions) can be represented as a sequence of generalized syntagmas that reflect the grammar properties of specific text words. On the basis of this concept the source sentence can be transformed into a set of templates of syntactic constructions accompanied by their “joint assemblies.” Then, the translation process can be performed by “disassembling” the source sentence into templates of syntactic constructions, correlating these source templates with templates of their translated equivalents and final “assembly” of the translated sentence in accordance with “technological instructions” in the “assembly rules.” This article proposes mechanisms for the automatic formation of a template representation of the hierarchy of syntactic constructions of bilingual sentences.

Efficient Low-Resource Neural Machine Translation with Reread and Feedback Mechanism

How to utilize information sufficiently is a key problem in neural machine translation (NMT), which is effectively improved in rich-resource NMT by leveraging large-scale bilingual sentence pairs. However, for low-resource NMT, lack of bilingual sentence pairs results in poor translation performance; therefore, taking full advantage of global information in the encoding-decoding process is effective for low-resource NMT. In this article, we propose a novel reread-feedback NMT architecture (RFNMT) for using global information. Our architecture builds upon the improved sequence-to-sequence neural network and consists of a double-deck attention-based encoder-decoder framework. In our proposed architecture, the information generated by the first-pass encoding and decoding process flows to the second-pass encoding process for more sufficient parameters initialization and information use. Specifically, we first propose a “reread” mechanism to transfer the outputs of the first-pass encoder to the second-pass encoder, and then the output is used for the initialization of the second-pass encoder. Second, we propose a “feedback” mechanism that transfers the first-pass decoder’s outputs to a second-pass encoder via an important weight model and an improved gated recurrent unit (GRU). Experiments on multiple datasets show that our approach achieves significant improvements over state-of-the-art NMT systems, especially in low-resource settings.

Retrosynthesis with attention-based NMT model and chemical analysis of "wrong" predictions.

We consider retrosynthesis to be a machine translation problem. Accordingly, we apply an attention-based and completely data-driven model named Tensor2Tensor to a data set comprising approximately 50 000 diverse reactions extracted from the United States patent literature. The model significantly outperforms the seq2seq model (37.4%), with top-1 accuracy reaching 54.1%. We also offer a novel insight into the causes of grammatically invalid SMILES, and conduct a test in which experienced chemists select and analyze the “wrong” predictions that may be chemically plausible but differ from the ground truth. The effectiveness of our model is found to be underestimated and the “true” top-1 accuracy reaches as high as 64.6%.

Khasi to English Neural Machine Translation: an Implementation Perspective

Being able to translate and communicate consistently from one language to another would have been the ultimate goal of an intelligent system. With recent advancement of Neural Machine Translation (NMT), it has shown a promising solution to the problem of machine translation. NMT generally requires large size parallel corpora to obtained a good translation accuracy. In this paper, we would like to explore a Translation system from Khasi to English language using both supervised and unsupervised technique. Unsupervised was inspired to help attaining a better translation accuracy for low resource language. It was influenced by the recent advancement of unsupervised neural machine translation which primarily relies on monolingual corpora. In this work, Supervised NMT technique was also implemented and compared with the standard OpenNMT toolkit. Here, we also use Statistical Machine Translation (SMT) tools like Moses as a standard benchmark to compare the translation accuracy. When considering monolingual corpus, we obtain an accuracy of 0.23%. Given the small size monolingual corpus the result was lacking but showed promising rooms for improvement. We obtain much better accuracy of 35.35% and 41.87% when we use parallel corpus in supervised NMT and OpenNMT respectively. On comparison with SMT system with Blue score of 43.76%, the supervised NMT system was on par in its performance. Lastly, with improvement in corpus size and better adaptation of preprocessing steps on the source language (Khasi) the result can be tune to a better outcome.