Hybrid Machine Translation Research Articles

Sanskrit to Hindi language translation using multimodal neural machine translation

Machine translation (MT) is a subfield of computer features that focuses on the automatic translation from one natural language into another without any human involvement. Due to native people interacting in a variety of languages, there is a great need for translating information between languages to send and communicate thoughts. However, they disregard the significance of semantic data encoded in the text features. In this paper, multimodal neural machine translation (MNMT) is proposed for Sanskrit-Hindi translation. The main goal of the proposed method is to fully utilize semantic text features on NMT architecture and to minimize testing and training time. The MNMT is validated on two different NMT architectures: recurrent neural network (RNN) and self-attention network (SAN). The MNMT method’s efficacy is demonstrated by employing the dataset of Sanskrit-Hindi Corpora. Extensive experimental outcomes represent the proposed method’s enhancement over baselines on both architectures. The existing methods, namely, English-to-Indian MT system, Sanskrit-Hindi MT system, and hybrid MT system are used to justify the efficacy of the MNMT method. When compared to the above-mentioned existing methods, RA-RNN respectively achieves a superior BLEU and METEOR of 80.5% and 75.3%, while the RA-SAN respectively achieves a superior BLEU and METEOR of 78.2% and 77.1%.

Implications of using AI in Translation Studies: Trends, Challenges, and Future Direction

This review paper provides an overview of the use of artificial intelligence (AI) in translation studies. The paper discusses the various AI techniques that have been used in translation, including statistical machine translation, rule-based machine translation, neural machine translation, and hybrid machine translation. The paper also explores the advantages and limitations of each model, as well as their applications in translation studies. Additionally, the paper reviews the various techniques for evaluating the effectiveness of AI models in translation and their advantages and limitations. The challenges and limitations of AI in translation, such as the handling of idioms, metaphors, and cultural nuances, are also discussed, along with research directions for improving AI-based translation. The review also discusses the representation of AI in literature and the arts, delves into the academic opportunities, and investigates its impact on human lives. Furthermore, AI's ethical and social implications in translation, such as job displacement, data privacy, and bias and fairness, are examined. Finally, the paper summarizes the main findings, implications, and recommendations for future research directions in AI-based translation studies. Overall, this review paper provides some insights into the current state of AI in translation and its potential for improving the field.

Experimental study of the impact of using hybride machine translation engines on the quality of translation of texts in the field of ornamental horticulture

The article is devoted to the study of the impact of using hybrid machine translation engines (Onlinedoctranslator and DeepL Translate) on the quality of translation of texts in the field of ornamental horticulture. Within our experimental study the quality of translation was determined by the number of errors and quality of rendering horticultural terminology. Machine translation is gradually evolving, the market is extremely active and proposes many new engines uniting most promising algorithms, but it is important to determine the effectiveness of hybrid machine translation engines in translating texts of various fields, especially in the field of ornamental horticulture. In the course of conducting our research, we formulated a research hypothesis, selected a text in the field of ornamental horticulture rich in horticultural terminology and consisting of more than 30.000 printed characters without spaces, compiled a bilingual glossary based on the selected text, consisting of 200 terminological units, selected machine translation engines (Onlinedoctranslator and DeepL) and the procedure of evaluating errors, having distinguished three main types of mistakes, analyzed translated texts in terms of general quality and in terms of transferring horticultural terminology, processed the experimental data, represented the results in the form of tables providing an expert assessment of the empirical data and formulated conclusions. In the process of conducting the experiment, the hypothesis formulated at the beginning of the research was confirmed: different hybrid machine translation engines offer translations of the same text in the field of ornamental horticulture of varying quality, which we tracked by determining the overall percentage of errors in each translation performed, as well as by calculating the percentage of correct transferring of horticultural terminology. According to both indicators, DeepL showed the best results, and therefore it is advisable to recommend it for translating texts in the field of ornamental horticulture.

Hybrid neural machine translation with statistical and rule based approach for syntactics and semantics between Tolaki-Indonesian-English languages

Hybrid neural machine translation with statistical and rule based approach for syntactics and semantics between Tolaki-Indonesian-English languages

Sentiment induced phrase-based machine translation: Robustness analysis of PBSMT with senti-module

Every type of machine translation system (i.e. neural, statistical, rule-based machine translation system) is equal important to build a sophistical hybrid machine translation system. Keeping this fact in my mind, I concentrate to improve statistical machine translation system with more natural way. In this paper, I try to preserve sentiment after translation to improve the overall accuracy of the machine translation system. So, I introduced senti-model here. A senti-model (sentiment model), translation model, language model, and distortion model are incorporated on the top of the beam search algorithm for decoding. At first, sentiment information is learned and modeled with translation probability by using this algorithm. Thereafter, I decode the source sentences-based on the contextual information. Overall procedure of translation modeling with a sentiment, parameter estimation for it, and senti-translation decoding (decoding with the sentiment model) are presented with empirical evidence. Experiments on a benchmark English–Hindi dataset shows that the proposed model is capable to improve the accuracy (in terms of 4.66 BLEU points, 4.09 LeBleu points, 4.67 NIST points, 5.71 RIBES points) significantly and preserves sentiment 7.79% more than the state-of-the-art technique.

Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine Translation

This paper proposes a hybrid machine translation (HMT) system that improves the quality of neural machine translation (NMT) by incorporating statistical machine translation (SMT). Therefore, two NMT systems and two SMT systems were built for the Slovenian–English language pair, each for translation in one direction. We used a multilingual language model to embed the source sentence and translations into the same vector space. From each vector, we extracted features based on the distances and similarities calculated between the source sentence and the NMT translation, and between the source sentence and the SMT translation. To select the best possible translation, we used several well-known classifiers to predict which translation system generated a better translation of the source sentence. The proposed method of combining SMT and NMT in the hybrid system is novel. Our framework is language-independent and can be applied to other languages supported by the multilingual language model. Our experiment involved empirical applications. We compared the performance of the classifiers, and the results demonstrate that our proposed HMT system achieved notable improvements in the BLEU score, with an increase of 1.5 points and 10.9 points for both translation directions, respectively.

Translation of English Language into Urdu Language Using LSTM Model

English to Urdu machine translation is still in its beginning and lacks simple translation methods to provide motivating and adequate English to Urdu translation. In order to make knowledge available to the masses, there should be mechanisms and tools in place to make things understandable by translating from source language to target language in an automated fashion. Machine translation has achieved this goal with encouraging results. When decoding the source text into the target language, the translator checks all the characteristics of the text. To achieve machine translation, rule-based, computational, hybrid and neural machine translation approaches have been proposed to automate the work. In this research work, a neural machine translation approach is employed to translate English text into Urdu. Long Short Term Short Model (LSTM) Encoder Decoder is used to translate English to Urdu. The various steps required to perform translation tasks include preprocessing, tokenization, grammar and sentence structure analysis, word embeddings, training data preparation, encoder-decoder models, and output text generation. The results show that the model used in the research work shows better performance in translation. The results were evaluated using bilingual research metrics and showed that the test and training data yielded the highest score sequences with an effective length of ten (10).

Machine Translation of English Language Using the Complexity-Reduced Transformer Model

Previous translation models like statistical machine translation (SMT), rule-based machine translation (RBMT), hybrid machine translation (HMT), and neural machine translation (NMT) have reached their performance bottleneck. The new Transformer-based machine translation model has become the favorite choice for English language translation. For instance, Google’s BERT translation model organizes the Transformer module into bidirectional encoder representations. It is aware of the users’ search intentions as well as the material that the search engine has indexed. It does not need to evaluate previous searches to comprehend what people mean, unlike RankBrain. BERT comprehends words, sentences, and complete information in the same way that we do. It achieves remarkable translation quality improvement over the other state-of-the-art benchmarks. It demonstrates the great potential of the Transformer model. The Transformer-based translation model mainly improves the performance at the cost of growing model sizes and complexity, usually requiring million-scale parameters. It is hard for the traditional computing systems to cope with the growing memory and computation requirements. However, the latest computers can easily run this model without any lag. The biggest challenge of applying the Transformer model is to deploy these models efficiently onto real-time or embedded devices. In this work, we propose a quantization scheme to reduce the parameter and computation complexity. It is of great importance to promote the usage of the Transformer model. Our experiment results show that the original Transformer model in 32 bit floating-point can be quantized to only 8 bits to 12 bits with only negligible translation quality loss. However, due to the perfect transformation of the block part, this quality loss part can easily be managed by the users. Meanwhile, our algorithm achieves 2.6 × to 4.0 × compression ratio, which is helpful to save the required complexity and energy during the inference phase.

Scaling up a Hybrid MT System: From low to full resources

This article describes a hybrid approach to machine translation (MT) that is inspired by the rule-based, statistical, example-based, and other hybrid machine translation approaches currently used or described in academic literature. It describes how the approach was implemented for language pairs using only limited monolingual resources and hardly any parallel resources (the METIS-II system), and how it is currently implemented with rich resources on both the source and target side as well as rich parallel data (the PaCo-MT system). We aim to illustrate that a similar paradigm can be used, irrespectively of the resources available, but of course with an impact on translation quality.

ОПРЕДЕЛЕНИЕ НАБОРА УСЛОВИЙ ДЛЯ АВТОМАТИЧЕСКОГО ПОИСКА ОПТИМАЛЬНОГО ВАРИАНТА ГИБРИДНОГО МАШИННОГО ПЕРЕВОДА ТЕКСТА НА ГРАФЕМНОМ УРОВНЕ

The article is devoted to the Algorithmic Search for Optimal Solutions for evaluating andimproving the Quality of Hybrid Machine Translation of Text. The Object of the Research is Textson any Alphabetical Languages with different Bases (Alphabets), as well as their Translations intoother Alphabetical Languages.Currently, existing Methods and Means of Hybrid Machine Translationare characterized by a wide variety of Quality Assessment Algorithms, but the Disadvantageof these Methods is that most of them do not have Clear Criteria, Limitations and Schemes of Assessments,eventually, the Result of the Translation in most cases does not correspond to the Levelof Publication. The Aim of the Work is to determine a Set of Conditions for automatic search forthe Best Option of Hybrid Machine Translation of Text at the Level of Graphemes.The Main Tasksto be solved during the Research are the Search for Qualitative and Quantitative Conditions, includingthe maximum, minimum and average values of the Lengths of Translations, Reverse Translationsand Editorial Distances between Pairs of Texts that have the Same Meaning. The ScientificNovelty lies in use the Graphical Representation of the Model of Alphabetic Languages at theLevel of Graphemes in the Form of a Cartesian Coordinate System with a Dimension equal to aUnit Editorial Distance (by Levenstein). When solving the de Goui’s Theorem, the current Rules ofStandardization PR 50.1.027–2014 "Rules for the Provision of Translation and Special Types ofLinguistic Services", the Method of Decanonicalization and the Model "Original Text – Translation– Reverse Translation" were used. As a Result, Actual and Practically Applicable Solutionsfor the Problems under consideration are obtained. In this regard; this Work may be interest to awide range of Specialists engaged in Machine Translation and Translation Studies.

Adaptation in Statistical Machine Translation for Low-resource Domains in English-Vietnamese Language

Adaptation in Statistical Machine Translation for Low-resource Domains in English-Vietnamese Language

Tourist ’s Translator based on Digital Image Processing and Hybrid Translation

Being a tourist in a foreign country is not easy when it comes to getting familiar with the local language. From reading signboards to getting unfairly charged while shopping, booking cabs and hotels for their stay, roaming around for sightseeing, communicating with the locals, everything requires an understanding of the local language. Nowadays, everybody has a smartphone, which proves to be the most helpful tool for travelling purposes. We aim to build up an Android application that will be capable of translating text, voice, and also textual information written on signboards from native language to the desired language, users have to use their smartphone camera for signboard translation. We are using a hybrid machine translation approach for language translation. Text recognition from the captured images of signboards is done with the help of digital image processing. The purpose of this work is to reduce the failure of any single machine translation approach and to reduce miscommunication between tourists and the local people.

A Survey on Hybrid Machine Translation

Machine translation has gradually developed in past 1940’s.It has gained more and more attention because of effective and efficient nature. As it makes the translation automatically without the involvement of human efforts. The distinct models of machine translation along with “Neural Machine Translation (NMT)” is summarized in this paper. Researchers have previously done lots of work on Machine Translation techniques and their evaluation techniques. Thus, we want to demonstrate an analysis of the existing techniques for machine translation including Neural Machine translation, their differences and the translation tools associated with them. Now-a-days the combination of two Machine Translation systems has the full advantage of using features from both the systems which attracts in the domain of natural language processing. So, the paper also includes the literature survey of the Hybrid Machine Translation (HMT).

Evaluation of the impact of controlled language on neural machine translation compared to other MT architectures

Many studies have shown that the application of controlled languages (CL) is an effective pre-editing technique to improve machine translation (MT) output. In this paper, we investigate whether this also holds true for neural machine translation (NMT). We compare the impact of applying nine CL rules on the quality of NMT output as opposed to that of rule-based, statistical, and hybrid MT by applying three methods: error annotation, human evaluation, and automatic evaluation. The analyzed data is a German corpus-based test suite of technical texts that have been translated into English by five MT systems (a neural, a rule-based, a statistical, and two hybrid MT systems). The comparison is conducted in terms of several quantitative parameters (number of errors, error types, quality ratings, and automatic evaluation metrics scores). The results show that CL rules positively affect rule-based, statistical, and hybrid MT systems. However, CL does not improve the results of the NMT system. The output of the neural system is mostly error-free both before and after CL application and has the highest quality in both scenarios among the analyzed MT systems showing a decrease in quality after applying the CL rules. The qualitative discussion of the NMT output sheds light on the problems that CL causes for this kind of MT architecture.

Hybrid Machine Translation by Combining Output from Multiple Machine Translation Systems

Hybrid Machine Translation by Combining Output from Multiple Machine Translation Systems

English-Arabic Hybrid Machine Translation System using EBMT and Translation Memory

The availability of a machine translation to translate from English-to-Arabic with high accuracy is not available because of the difficult morphology of the Arabic Language. A hybrid machine translation system between Example Based machine translation technique and Translation memory was introduced in this paper. Two datasets have been used in the experiments that were constructed by using internal medicine publications and Worldwide Arabic Medical Translation Guide Common Medical Terms sorted by Arabic. To examine the accuracy of the system constructed four experiments were made using Example Based Machine Translation system in the first, Google Translate in the second and Example Based with Google translate in the third and the fourth is the system proposed using Example Based with Translation memory. The system constructed achieved 77.17 score for the first dataset and 63.85 score for the second which were the highest score using BLEU score.

Pause-Based Phrase Extraction and Effective OOV Handling for Low-Resource Machine Translation Systems

Machine translation is the core problem for several natural language processing research across the globe. However, building a translation system involving low-resource languages remains a challenge with respect to statistical machine translation (SMT). This work proposes and studies the effect of a phrase-induced hybrid machine translation system for translation from English to Tamil, under a low-resource setting. Unlike conventional hybrid MT systems, the free-word ordering feature of the target language Tamil is exploited to form a re-ordered target language model and to extend the parallel text corpus for training the SMT. In the current work, a novel rule-based phrase-extraction method, implemented using parts-of-speech (POS) and place-of-pause in both languages is proposed, which is used to pre-process the training corpus for developing the back-off phrase-induced SMT. Further, out-of-vocabulary (OOV) words are handled using speech-based transliteration and two-level thesaurus intersection techniques based on the POS tag of the OOV word. To ensure that the input with OOV words does not skip phrase-level translation in the hierarchical model, a phrase-level example-based machine translation approach is adopted to find the closest matching phrase and perform translation followed by OOV replacement. The proposed system results in a bilingual evaluation understudy score of 84.78 and a translation edit rate of 19.12. The performance of the system is compared in terms of adequacy and fluency, with existing translation systems for this specific language pair, and it is observed that the proposed system outperforms its counterparts.

Hybrid Machine Translation with Multi-Source Encoder-Decoder Long Short-Term Memory in English-Malay Translation

<p class='IJASEITAbtract'>Statistical Machine Translation (SMT) and Neural Machine Translation (NMT) are the state-of-the-art approaches in machine translation (MT). The translation produced by a SMT is based on the statistical analysis of text corpora, while NMT uses deep neural network to model and to generate a translation. SMT and NMT have their strength and weaknesses. SMT may produce better translation with a small parallel text corpus compared to NMT. Nevertheless, when the amount of parallel text available is large, the quality of the translation produced by NMT is often higher than SMT. Besides that, study also shown that the translation produced by SMT is better than NMT in cases where there is a domain mismatch between training and testing. SMT also has an advantage on long sentences. In addition, when a translation produced by an NMT is wrong, it is very difficult to find the error. In this paper, we investigate a hybrid approach that combine SMT and NMT to perform English to Malay translation. The motivation of using a hybrid machine translation is to combine the strength of both approaches to produce a more accurate translation. Our approach uses the multi-source encoder-decoder long short-term memory (LSTM) architecture. The architecture uses two encoders, one to embed the sentence to be translated, and another encoder to embed the initial translation produced by SMT. The translation from the SMT can be viewed as a “suggestion translation” to the neural MT. Our experiments show that the hybrid MT increases the BLEU scores of our best baseline machine translation in computer science domain and news domain from 21.21 and 48.35 to 35.97 and 61.81 respectively.

An online English-Khmer hybrid machine translation system

The present paper describes design of an online hybrid machine translation (MT) system involving a low-resource language Khmer, the official language of Cambodia. The proposed system uses an open source statistical machine translation (SMT) toolkit DoMY CE as the primary translation tool. The parallel corpora have been prepared from various sources and the Headley Khmer-English dictionary. Language model, translation model and decoder configurations have been done using the DOMY toolkit. We used a post-processing step of using parts of speech tagger to enhance the quality of target language sentence. Experimental results demonstrate the success of the proposed scheme with English as source and Khmer as the target language. In our experiments the proposed model achieved significantly good National Institute of Standards and Technology (NIST) and BiLingual Evaluation Understudy (BLEU) scores. Different web technologies have been used for developing an online translation system.

An online English-Khmer hybrid machine translation system

The present paper describes design of an online hybrid machine translation (MT) system involving a low-resource language Khmer, the official language of Cambodia. The proposed system uses an open source statistical machine translation (SMT) toolkit DoMY CE as the primary translation tool. The parallel corpora have been prepared from various sources and the Headley Khmer-English dictionary. Language model, translation model and decoder configurations have been done using the DOMY toolkit. We used a post-processing step of using parts of speech tagger to enhance the quality of target language sentence. Experimental results demonstrate the success of the proposed scheme with English as source and Khmer as the target language. In our experiments the proposed model achieved significantly good National Institute of Standards and Technology (NIST) and BiLingual Evaluation Understudy (BLEU) scores. Different web technologies have been used for developing an online translation system.