Machine Translation Research Articles

Improving neural machine translation by integrating transliteration for low-resource English–Assamese language

Improving neural machine translation by integrating transliteration for low-resource English–Assamese language

Google and Legal Translation: The Case Study of Contracts

In recent years, the need for Machine Translation (MT) has grown, especially for translating legal contracts between languages like Arabic and English. This study primarily investigates whether Google Translator can adequately replace human translation for legal documents. Utilizing a widely popular free web-based tool, Google Translate, the research method involved translating six segments from various legal contracts into Arabic and assessing the translations for lexical and syntactic accuracy. The findings show that although Google Translate can quickly produce English-Arabic translations, it falls short compared to professional translators, especially with complex legal terms and syntax. Errors can be categorized into: polysemy, homonymy, legal doublets, and adverbs at the linguistic level, and morphological parsing, concord, and modality at the syntactic level. The study concludes with recommendations for enhancing machine translation systems and suggests caution in using Google Translate for legal purposes, advocating for continued reliance on human expertise in legal settings.

Foreword

The diversity of interdisciplinary research has encouraged scholars in the Arab region and in other parts of the world to explore the intersectionality of different disciplines. This exploration of intersectionality requires researchers to move away from traditional approaches and seek innovative ways to address areas of research that may have not yet been investigated or have emerged as new trends. Such endeavors enrich our understanding of existing knowledge and build upon existing information. The current AWEJ for Translation and Literary Studies issue has attracted works from researchers across disciplines and invites readers, both specialized and nonspecialized, to come abreast with new insights around machine translation, the quality of digital communication, crime against children, gender differences, vegetarian ecofeminism, masculine food studies, humanity, the notion of identity, linguistics and standup comedy, emerging Arab women literature, the configuration of the social and cultural aspects of freedom, the restrictions of confinement and the sense of belonging amidst the chaos of war and revolutions. Such a wide spectrum of research work is a reflection of the growing influence of digital technologies in creating opportunities for innovation and the development of new scopes of knowledge. The very concept of cross disciplinary approaches challenges researchers to tap into the grey areas and create novel connections which address both regional and global issues. Encouraging interdisciplinarity stems from the deep conviction that the world today needs unique approaches to understanding current shared challenges such as the frailty of communication between communities, security and safety amidst wars and hunger, identity against the backdrop of political uncertainty, food security, water crisis, and humanitarian crisis. Research that addresses and advances progressive thought processes with groundbreaking insights and or impact is the basis of change. With this in mind, it remains the duty of journals and reviewers to join efforts to create the right basis for quality research and impactful results to emerge

Talking Hands: Speech to Sign Language Converter

Deaf people face communication challenges in many aspects of their lives. This project proposes a speech-to-sign language converter that can help to bridge the communication gap between deaf and hearing people. The converter would use speech recognition technology to transcribe spoken language into text. This text would then be translated into sign language images or GIFs. The images or GIFs would be displayed on a screen, allowing deaf people to easily understand the message. The converter would be designed to be as accurate and user-friendly as possible. It would also be customizable to support different sign languages. This project has the potential to make a significant impact on the lives of deaf people. It could help them to communicate more effectively in school, at work, and in social settings. It could also help to break down communication barriers and foster greater understanding between deaf and hearing people. The converter would be developed using a variety of technologies, including speech recognition, machine translation, and image processing. The objective of our project is to reduce the communication gap between normal people and disabled people by providing a virtual animation translator for sign language.

E-learning application in english writing classroom based on neural machine translation and semantic analysis algorithms

At present, there are some problems in traditional English writing teaching, such as the lack of practice opportunities and the lack of timely feedback. The aim of this study is to design an e-learning platform to improve students’ writing ability through neural machine translation and semantic analysis algorithms. This paper introduces the basic principles of neural machine translation and semantic analysis algorithms and their applications in translation and semantic understanding. In this paper, a computer English writing correction model based on neural machine translation and semantic analysis algorithm is proposed and applied to e-learning platform. The platform structure is constructed. The neural machine translation model is used to translate and transform the original writing content into standard English expressions. Then, semantic analysis algorithm is used to detect errors and propose suggestions for modification. The results show that the e-learning platform can provide personalized and timely feedback, and students’ writing ability has been significantly improved on the e-learning platform. They have improved their writing skills and presentation through exercises and submissions through the e-learning platform.

Enhancing Post-Editing Machine Translation Skills Among Iraqi Undergraduate Students

Post-editing machine translation often results in discrepancies between industry expectations and the actual quality of the final product. Despite the usually, low-quality outcomes associated with machine translation, it remains underutilized, even when there are opportunities for improvement through post-editing jobs. This study focuses on the impact of technology-assisted training on English proficiency among students. It investigates a standard error observed among Iraqi students during the post-editing of machine translations using Google Translate deciding whether to edit the machine translation output or to translate from scratch and understand the source text. This research involved 20 undergraduates from Al-Mustansiriyah University and employed a mixed methods approach, collecting data through an online questionnaire completed by students in the undergraduate translation program. The findings indicate a significant improvement in the frequency of post-editing skills among Iraqi students, serving as a distinguishing factor between the two groups. It also highlighted differences in translation abilities among students in Iraqi higher education, with notable improvements observed in the experimental group. In summary, this study supports previous research from the last five years, indicating a more favourable attitude toward machine translation and suggesting that post-editing is considered an essential sub-competency in translation education.

CodonBERT: a BERT-based architecture tailored for codon optimization using the cross-attention mechanism.

Due to the varying delivery methods of mRNA vaccines, codon optimization plays a critical role in vaccine design to improve the stability and expression of proteins in specific tissues. Considering the many-to-one relationship between synonymous codons and amino acids, the number of mRNA sequences encoding the same amino acid sequence could be enormous. Finding stable and highly expressed mRNA sequences from the vast sequence space using in silico methods can generally be viewed as a path-search problem or a machine translation problem. However, current deep learning-based methods inspired by machine translation may have some limitations, such as recurrent neural networks, which have a weak ability to capture the long-term dependencies of codon preferences. We develop a BERT-based architecture that uses the cross-attention mechanism for codon optimization. In CodonBERT, the codon sequence is randomly masked with each codon serving as a key and a value. In the meantime, the amino acid sequence is used as the query. CodonBERT was trained on high-expression transcripts from Human Protein Atlas mixed with different proportions of high codon adaptation index codon sequences. The result showed that CodonBERT can effectively capture the long-term dependencies between codons and amino acids, suggesting that it can be used as a customized training framework for specific optimization targets. CodonBERT is freely available on https://github.com/FPPGroup/CodonBERT.

Stameering Speech Signal Segmentation and Classification using Machine Learning

Stuttering or Stammering is a speech defect within which sounds, syllables, or words are rehashed or delayed, disrupting the traditional flow of speech. Stuttering can make it hard to speak with other individuals, which regularly have an effect on an individual's quality of life. Automatic Speech Recognition (ASR) system is a technology that converts audio speech signal into corresponding text. Presently ASR systems play a major role in controlling or providing inputs to the various applications. Such an ASR system and Machine Translation Application suffers a lot due to stuttering (speech dysfluency). Dysfluencies will affect the phrase consciousness accuracy of an ASR, with the aid of increasing word addition, substitution and dismissal rates. In this work we focused on detecting and removing the prolongation, silent pauses and repetition to generate proper text sequence for the given stuttered speech signal. The stuttered speech recognition consists of two stages namely classification using ANN and testing in ASR. The major phases of classification system are Re-sampling, Segmentation, Pre Emphasis, Epoch Extraction and Classification. The current work is carried out in UCLASS Stuttering dataset using MATLAB with 4% to 6% increase in accuracy by ANN.

Breaking language barriers with ChatGPT: enhancing low-resource machine translation between algerian arabic and MSA

Breaking language barriers with ChatGPT: enhancing low-resource machine translation between algerian arabic and MSA

Desktop Assistant Based on NLP

Natural Language Processing (NLP) has emerged as a critical component of artificial intelligence, enabling machines to comprehend and interact with human language. This research paper explores the current state of the art in NLP, highlighting recent innovations, trends, and ongoing challenges. It delves into various applications of NLP, discusses the datasets and models that drive advancements, and examines the evaluation metrics used to assess NLP systems. Key innovations such as transformers, pre-trained language models, and transfer learning have revolutionized the field, leading to significant improvements in performance across a variety of tasks. Additionally, the paper addresses the growing emphasis on ethical AI and bias mitigation, as well as the integration of NLP with other AI technologies to create multimodal systems. Applications of NLP in text classification, sentiment analysis, machine translation, conversational agents, and information retrieval are thoroughly examined. The discussion extends to the critical role of benchmark datasets and pre-trained models in driving progress. Furthermore, the paper evaluates the effectiveness of various metrics used to measure the performance of NLP systems. Finally, the future prospects and potential research directions are considered, highlighting the ongoing efforts to push the boundaries of what NLP can achieve in an increasingly interconnected and data-driven world. Keywords Natural language processing . Natural language understanding . Natural language generation

Generative design of compounds with desired potency from target protein sequences using a multimodal biochemical language model

Deep learning models adapted from natural language processing offer new opportunities for the prediction of active compounds via machine translation of sequential molecular data representations. For example, chemical language models are often derived for compound string transformation. Moreover, given the principal versatility of language models for translating different types of textual representations, off-the-beaten-path design tasks might be explored. In this work, we have investigated generative design of active compounds with desired potency from target sequence embeddings, representing a rather provoking prediction task. Therefore, a dual-component conditional language model was designed for learning from multimodal data. It comprised a protein language model component for generating target sequence embeddings and a conditional transformer for predicting new active compounds with desired potency. To this end, the designated “biochemical” language model was trained to learn mappings of combined protein sequence and compound potency value embeddings to corresponding compounds, fine-tuned on individual activity classes not encountered during model derivation, and evaluated on compound test sets that were structurally distinct from training sets. The biochemical language model correctly reproduced known compounds with different potency for all activity classes, providing proof-of-concept for the approach. Furthermore, the conditional model consistently reproduced larger numbers of known compounds as well as more potent compounds than an unconditional model, revealing a substantial effect of potency conditioning. The biochemical language model also generated structurally diverse candidate compounds departing from both fine-tuning and test compounds. Overall, generative compound design based on potency value-conditioned target sequence embeddings yielded promising results, rendering the approach attractive for further exploration and practical applications.Scientific contributionThe approach introduced herein combines protein language model and chemical language model components, representing an advanced architecture, and is the first methodology for predicting compounds with desired potency from conditioned protein sequence data.

Is machine interpreting interpreting?

Abstract This article first considers the question whether machine translation is translation and moves on to address the analogous issue for interpreting. After a review of the development and state of the art in machine interpreting, more commonly referred to as ‘spoken language translation’ or ‘speech translation’, the question of whether machine interpreting is interpreting is discussed – first with regard to terminology and conceptual distinctions and then in broader translation-theoretical frameworks. Using Otto Kade’s early definitional proposal as a point of departure, a reconceptualization is proposed in the form of a three-dimensional model designed to go beyond rigid taxonomies. The dimensions of agency, embodiment and immediacy are used to characterize translation as a graded concept in which these features may be more or less prominent.

Artificial Intelligence Impact on Human Translation: Legal Texts as a Case Study

The recent paper highlights the impact of artificial intelligence on Machine Translation without the interaction of Humans. The use of Google Translator, Bing, Microsoft Translator, Systran Translate and Amazon Translate has become widely spread (CAT Tools). This study aims to reveal the contrast between Artificial Intelligence and Human Translation in the legal field. A hypothesis of the difference between Artificial Translation and Human Translation was raised. The concerns about the lack of a translator increased, and machine translation was selected as the most selected option. Local and foreign contracts were selected and subjected to Human and Machine translation. Strengths and weaknesses points were selected and analyzed. The previous studies in the legal translation field were considered. The results revealed the gap between human translation and machine translation, and human translation is dominant in the light of accuracy and the existence of legal language. The findings also focused on the Translators' experience and knowledge in the translation field.

THÌ and LÀ Constructions in Vietnamese: A Study on Argument Structures

This study aims to shed light on the argument structure of the Vietnamese particles THÌ and LÀ, which function as markers of focused constituents and topic-comment structures, respectively. The research objectives are to identify the syntactic and semantic properties of constructions involving THÌ and LÀ, and to analyze their argument structures within the theoretical frameworks of construction grammar. Specifically, the constructional approach (Goldberg 1995, 2006) and radical construction grammar (Croft 2001) will be employed. The methodology involves collecting a corpus of naturally occurring data from various written sources. The data will undergo qualitative analysis to identify argument structure patterns associated with THÌ and LÀ constructions, as well as quantitative analysis of frequency and distribution. The expected findings will provide insights into the argument structures of these Vietnamese constructions, elucidating how focused constituents and topic-comment structures are marked. The study will evaluate the applicability of the chosen theoretical frameworks in accounting for the observed patterns. The findings have potential applications in natural language processing tasks requiring accurate recognition and interpretation of focused constituents and topic-comment structures, such as information extraction, machine translation, and language generation. Furthermore, the study may inform language teaching and acquisition by shedding light on the usage patterns of THÌ and LÀ in Vietnamese. The analysis is expected to contribute to a deeper understanding of argument structures in Vietnamese and potentially motivate extensions or modifications to the theoretical frameworks.

Towards Media Monitoring: Detecting Known and Emerging Topics through Multilingual and Crosslingual Text Classification

This study aims to address challenges in media monitoring by enhancing closed-set topic classification in multilingual contexts (where both training and testing occur in several languages) and crosslingual contexts (where training is in English and testing spans all languages). To achieve this goal, we utilized a dataset from the European Media Monitoring webpage, which includes approximately 15,000 article titles across 18 topics in 58 different languages spanning a period of nine months from May 2022 to March 2023. Our research conducted comprehensive comparative analyses of nine approaches, encompassing a spectrum of embedding techniques (word, sentence, and contextual representations) and classifiers (trainable/fine-tunable, memory-based, and generative). Our findings reveal that the LaBSE+FFNN approach achieved the best performance, reaching macro-averaged F1-scores of 0.944 ± 0.015 and 0.946 ± 0.019 in both multilingual and crosslingual scenarios. LaBSE+FFNN’s similar performance in multilingual and crosslingual scenarios eliminates the need for machine translation into English. We also tackled the open-set topic classification problem by training a binary classifier capable of distinguishing between known and new topics with the average loss of ∼0.0017 ± 0.0002. Various feature types were investigated, reaffirming the robustness of LaBSE vectorization. The experiments demonstrate that, depending on the topic, new topics can be identified with accuracies above ∼0.796 and of ∼0.9 on average. Both closed-set and open-set topic classification modules, along with additional mechanisms for clustering new topics to organize and label them, are integrated into our media monitoring system, which is now used by our real client.

A Brief Review on Preprocessing Text in Arabic Language Dataset: Techniques and Challenges

Text preprocessing plays an important role in natural language processing (NLP) tasks containing text classification, sentiment analysis, and machine translation. The preprocessing of Arabic text still presents unique challenges due to the language's rich morphology, complex grammar, and various character sets. This brief review studied various techniques utilized for preprocessing Arabic text data. This study discusses the challenges specific to Arabic text and current an overview of key preprocessing steps including normalization, tokenization, stemming, stop-word removal, and noise reduction. This survey analyzes preprocessing techniques on NLP tasks and focus on current research trends and future directions in Arabic text preprocessing.

A survey of context in neural machine translation and its evaluation

Abstract The question of context in neural machine translation often focuses on topics related to document-level translation or intersentential context. However, there is a wide range of other aspects that can be considered under the umbrella of context. In this work, we survey ways that researchers have incorporated context into neural machine translation systems and the evaluation thereof. This includes building translation systems that operate at the paragraph level or the document level or ones that translate at the sentence level but incorporate information from other sentences. We also consider how issues like terminology consistency, anaphora, and world knowledge or external information can be considered as types of context relevant to the task of machine translation and its evaluation. Closely tied to these topics is the question of how to best evaluate machine translation output in a way that is sensitive to the contexts in which it appears. To this end, we discuss work on incorporating context into both human and automatic evaluations of machine translation quality. Furthermore, we also discuss recent experiments in the field as they relate to the use of large language models in translation and evaluation. We conclude with a view of the future of machine translation, where we expect to see issues of context continue to come to the forefront.

Neural Machine Translation for Low-Resource Languages from a Chinese-centric Perspective: A Survey

Machine translation—the automatic transformation of one natural language (source language) into another (target language) through computational means—occupies a central role in computational linguistics and stands as a cornerstone of research within the field of Natural Language Processing (NLP). In recent years, the prominence of Neural Machine Translation (NMT) has grown exponentially, offering an advanced framework for machine translation research. It is noted for its superior translation performance, especially when tackling the challenges posed by low-resource language pairs that suffer from a limited corpus of data resources. This article offers an exhaustive exploration of the historical trajectory and advancements in NMT, accompanied by an analysis of the underlying foundational concepts. It subsequently provides a concise demarcation of the unique characteristics associated with low-resource languages and presents a succinct review of pertinent translation models and their applications, specifically within the context of languages with low-resources. Moreover, this article delves deeply into machine translation techniques, highlighting approaches tailored for Chinese-centric low-resource languages. Ultimately, it anticipates upcoming research directions in the realm of low-resource language translation.

Ascle-A Python Natural Language Processing Toolkit for Medical Text Generation: Development and Evaluation Study.

Medical texts present significant domain-specific challenges, and manually curating these texts is a time-consuming and labor-intensive process. To address this, natural language processing (NLP) algorithms have been developed to automate text processing. In the biomedical field, various toolkits for text processing exist, which have greatly improved the efficiency of handling unstructured text. However, these existing toolkits tend to emphasize different perspectives, and none of them offer generation capabilities, leaving a significant gap in the current offerings. This study aims to describe the development and preliminary evaluation of Ascle. Ascle is tailored for biomedical researchers and clinical staff with an easy-to-use, all-in-one solution that requires minimal programming expertise. For the first time, Ascle provides 4 advanced and challenging generative functions: question-answering, text summarization, text simplification, and machine translation. In addition, Ascle integrates 12 essential NLP functions, along with query and search capabilities for clinical databases. We fine-tuned 32 domain-specific language models and evaluated them thoroughly on 27 established benchmarks. In addition, for the question-answering task, we developed a retrieval-augmented generation (RAG) framework for large language models that incorporated a medical knowledge graph with ranking techniques to enhance the reliability of generated answers. Additionally, we conducted a physician validation to assess the quality of generated content beyond automated metrics. The fine-tuned models and RAG framework consistently enhanced text generation tasks. For example, the fine-tuned models improved the machine translation task by 20.27 in terms of BLEU score. In the question-answering task, the RAG framework raised the ROUGE-L score by 18% over the vanilla models. Physician validation of generated answers showed high scores for readability (4.95/5) and relevancy (4.43/5), with a lower score for accuracy (3.90/5) and completeness (3.31/5). This study introduces the development and evaluation of Ascle, a user-friendly NLP toolkit designed for medical text generation. All code is publicly available through the Ascle GitHub repository. All fine-tuned language models can be accessed through Hugging Face.

Research on grammatical error correction algorithm in English translation via deep learning

Abstract This study provides a concise overview of a grammatical error correction algorithm that is based on an encoder-decoder machine translation structure. Additionally, it incorporates the attention mechanism to enhance the algorithm’s performance. Subsequently, simulation experiments were conducted to compare the improved algorithm with an algorithm based on a classification model and an algorithm based on the traditional translation model using open corpus data and English translations from freshmen. The results demonstrated that the optimized algorithm yielded superior intuitive error correction outcomes. When applied to both the open corpus and the English translations of college freshmen, the optimized error correction algorithm outperformed the others. The traditional translation model-based algorithm came in second, while the classification model-based algorithm showed the least favorable performance. Furthermore, all three error correction algorithms experienced a decrease in performance when dealing with English compositions from freshmen. However, the optimized algorithm exhibited a relatively smaller decline.