Comparative Study of Translation Effects in Low-Resource Languages Based on Pivot Language Method

Over the past few years, Large Language Models of Code (Code LLMs) have started to have a significant impact on programming practice. Code LLMs are also emerging as building blocks for research in programming languages and software engineering. However, the quality of code produced by a Code LLM varies significantly by programming language. Code LLMs produce impressive results on high-resource programming languages that are well represented in their training data (e.g., Java, Python, or JavaScript), but struggle with low-resource languages that have limited training data available (e.g., OCaml, Racket, and several others). This paper presents an effective approach for boosting the performance of Code LLMs on low-resource languages using semi-synthetic data. Our approach, called M ulti PL-T, generates high-quality datasets for low-resource languages, which can then be used to fine-tune any pretrained Code LLM. M ulti PL-T translates training data from high-resource languages into training data for low-resource languages in the following way. 1) We use a Code LLM to synthesize unit tests for commented code from a high-resource source language, filtering out faulty tests and code with low test coverage. 2) We use a Code LLM to translate the code from the high-resource source language to a target low-resource language. This gives us a corpus of candidate training data in the target language, but many of these translations are wrong. 3) We use a lightweight compiler to compile the test cases generated in (1) from the source language to the target language, which allows us to filter our obviously wrong translations. The result is a training corpus in the target low-resource language where all items have been validated with test cases. We apply this approach to generate tens of thousands of new, validated training items for five low-resource languages: Julia, Lua, OCaml, R, and Racket, using Python as the source high-resource language. Furthermore, we use an open Code LLM (StarCoderBase) with open training data (The Stack), which allows us to decontaminate benchmarks, train models without violating licenses, and run experiments that could not otherwise be done. Using datasets generated with M ulti PL-T, we present fine-tuned versions of StarCoderBase and Code Llama for Julia, Lua, OCaml, R, and Racket that outperform other fine-tunes of these base models on the natural language to code task. We also present Racket fine-tunes for two very recent models, DeepSeek Coder and StarCoder2, to show that M ulti PL-T continues to outperform other fine-tuning approaches for low-resource languages. The M ulti PL-T approach is easy to apply to new languages, and is significantly more efficient and effective than alternatives such as training longer.

This paper describes the extended pivot-based approach for bilingual lexicon extraction. The basic features of the approach can be described as follows: First, the approach builds context vectors between a source (or target) language and a pivot language like English, respectively. This is the same as the standard pivot-based approach which is useful for extracting bilingual lexicons between low-resource languages such as Korean-French. Second, unlike the standard pivot-based approach, the approach looks for similar context vec- tors in a source language. This is helpful to extract translation candidates for polysemous words as well as lets the translations be more confident. Third, the approach extracts translation candidates from target context vectors through the similarity between source and target context vectors. Based on these features, this paper describes the extended pivot-based approach and does various experiments in a language pair, Korean-French (KR-FR). We have observed that the approach is useful for extracting the most proper translation candidate as well as for a low-resource language pair.

This paper is an attempt to focus on investigating the pivot (bridge) language technique, where the pivot language used to improve Statistical Machine Translation (SMT) quality. In this case, Indonesian is used as a pivot language, where each available corpus can be used to support the Madurese-Sundanese language pair. Experiments that have been carried out using the parallel corpus of the Indonesian-Madurese and Indonesian-Sundanese languages ​​are equal to 5K and 6K sentences respectively, while the monolingual corpus used Malay, Sundanese and Indonesian each at 10K, 10K and 100K sentences. This study compares the results of applying the Triangulation and Transfer methods using Indonesian as a pivot language. The results of the research proved that the Triangulation method has better acceleration when compared to the Transfer method. From the results of the experiments conducted, the Triangulation method increased the average Indonesian pivot-based SMT testing by 6.18% for Madura-Sundanese SMT and 7.27% for Madurese-Sundanese SMT.

Multi-Language Speech-to-Text Translation (ST) plays a pivotal role in bridging linguistic barriers by converting spoken language into written text across different languages.This project aims to develop a robust ST model tailored for low-resource Indian languages, specifically targeting the Indo-Aryan and Dravidian language families.The dataset used consists of speeches from conferences and TED Talks, along with their corresponding transcriptions in English (source language) and translations in Hindi, Bengali, and Tamil (target languages).By tackling the lack of data and disparities in attention within low-resource languages, the paper strives to create an efficient ST system capable of real-world deployment.Additionally, existing resources in related languages are leveraged and word-level translation resources are explored to enhance translation accuracy.

Background/Objective: The shift from classical rule-based and statistical systems of Machine Translation (MT) to more advanced Neural Machine Translation (NMT) systems employing transformers is quite astonishing. Still, translating language pairs such as English and Amharic, which are structurally different, poses a considerable challenge. This study explores the impact of incorporating known syntactic information on the performance of English-Amharic translation in Graph2Seq model to the precision and fluency of translations. Methods: We aimed to enhance translation performance with the design of an attention-based Graph2Seq model that utilizes a shaped syntactic dependency tree and implements Graph Neural Networks (GNN) to capture the hierarchy of language. The model overcomes the constraints posed by conventional sequence-to-sequence (Seq2Seq) models by adding syntactic information into the encoder. The dataset includes roughly 1.14 million parallel sentences in English and Amharic. Findings: There is a positive, actionable gap, as the proposed model was able to exceed the benchmarks set by the transformer models and pretrained methodologies, attaining BLEU 37.3, and demonstrating marked improvement in translation quality. Furthermore, these findings have defied the argument stating that diversity on the morpho-syntactic level impedes translation, instead illustrating the potency of rich superset syntax modeling. This score beats baseline figures by 4.56 points and 24.24 points compared to M2M100 and the standard transformer models, respectively. Novelty: All these points towards the fact that incorporating deep syntactic structures to enrich Graph-NMT systems on low-resource languages enhances their capabilities, paving avenues for further investigations in the field of Machine Translation. This study can be a starting point towards more research on the enhancement of accuracy and fluency of translations using a multidisciplinary approach based on language. Keywords: Graph neural networks, English-Amharic machine translation, Syntactic of source language, BLEU score, Low-resource language

This study examines the efficacy of employing Indonesian as an intermediary language to improve the quality of translations from Javanese to Madurese through a pivot-based approach utilizing neural machine translation (NMT). The principal objective of this research is to enhance translation precision and uniformity among these low-resource languages, hence advancing machine translation models for underrepresented languages. The data collecting approach entailed extracting parallel texts from internet sources, followed by pre-processing through tokenization, normalization, and stop-word elimination algorithms. The prepared datasets were utilized to train and assess the NMT models. An intermediary phase utilizing Indonesian is implemented in the translation process to enhance the accuracy and consistency of translations between Javanese and Madurese. Parallel text corpora were created by collecting and preprocessing data, thereafter, utilized to train and assess the NMT models. The pivot-based strategy regularly surpassed direct translation regarding BLEU scores for all n-grams (BLEU-1 to BLEU-4). The enhanced BLEU ratings signify increased precision in vocabulary selection, preservation of context, and overall comprehensibility. This study significantly enhances the current literature in machine translation and computational linguistics, especially for low-resource languages, by illustrating the practical effectiveness of a pivot-based method for augmenting translation precision. The method's dependability and efficacy in producing genuine translations were proved through numerous studies. The pivot-based technique enhances translation quality, although it possesses limitations, including the risk of error propagation and bias originating from the pivot language. Further research is necessary to examine the integration of named entity recognition (NER) to improve accuracy and optimize the intermediate translation process. This project advances the domains of machine translation and the preservation of low-resource languages, with practical implications for multilingual communities, language education resources, and cultural conservation.

This study investigates the critical yet understudied influence of translation direction on both neural machine translation post-editing (PE) and human translation (HT) processes. While previous research has compared PE and HT, the interaction between translation mode and directionality-particularly for linguistically distant languages-remains poorly understood. We address this gap by examining how L1-L2 (Chinese-English) versus L2-L1 (English-Chinese) directionality affects translation performance across multiple dimensions. Thirty-one native Chinese translators completed both PE and HT tasks in both language directions, with their processes documented through keylogging and screen recording. Our findings reveal that PE generally outperforms HT in speed, technical efficiency, and output quality, but these advantages manifest differently depending on translation direction. The PE advantage was significantly more pronounced in the L2-L1 direction. Conversely, perceived task difficulty was primarily determined by directionality rather than translation mode, with L1-L2 translation consistently rated more challenging regardless of whether performed through PE or HT. These results demonstrate the complex interplay between translation mode and directionality, challenging the simplistic view that PE universally reduces effort compared to HT. Our findings have important implications for translator training, workflow optimization, and fair compensation models that account for both translation mode and direction.

This study evaluates the dual role of translators and interpreters working for NGOs involved in the health sector in developing countries. Given the crucial importance of public health in a nation’s development, we assess the way grassroots participation in healthcare projects and activities is enhanced through the translation of healthcare information into low-resource languages. This study, epistemologically grounded in the theory of planned behaviour (TPB), adopts a qualitative and quantitative methodological approach to correlate the TPB and the target-language choices of translator-interpreters and the translation strategies they adopt working with low-resource languages. The survey responses and audio recordings analysed are those of translators or interpreters working in a rural context in the Anglophone regions of Cameroon. The target language in question is Pidgin English, a low-resource oral language. The evidence gathered leads to the finding that various de-terminologization strategies are employed to ensure that messages are adequately disseminated and that the local community members are involved in healthcare activities. The study seeks to bring together translation, interpreting, the TPB, and international development in which the involvement of NGOs underscores North–South cooperation in the strategic domain of global healthcare.

This is the first review of the existing research on the topic of indirect translation in Dutch-Russian language pair. Both theoretical justifications for the study of this problem (H. Pięta, M. Ringmar, C. Dollerup, etc.) and the results of empirical research (I. E. Kuznetsova, H. van der Tak, I. M. Michajlova, etc.) are presented. The paper includes observations on the influence of the mediating language on the creation of the first dictionaries, translation of scientific works, legal documents and fiction, and separately highlights the issue of using English as a “pivot language” in popular online translators. The main problem in this case becomes the translation of personal pronouns and homonyms. In recent years, on the one hand, researchers have been fighting the stigmatisation of mediated translation as a priori inferior and proving that mediated languages link distant cultures and play an important role in the dissemination of literature. On the other hand, textual analyses of different types of indirect translations reveal errors and inaccuracies that could have been easily avoided in direct translation. An intermediary translation can be seen as a translation made with the help of a language other than the source language, but not necessarily from a “third language”: when translation plurality occurs, we can often speak of intralanguage indirect translation. This paper uses Multatuli’s novel Max Havelaar as an illustration, in particular the 1959 Russian edition, where German and Russian translations mediate. Resorting to the help of their predecessors, the Russian translators omit the same fragments, use similar paraphrases, and repeat lexical and phonetic distortions. The use of mediating texts is certainly related to extra-linguistic factors: lack of translators, demand for texts, publishers’ desire to reduce costs or speed up the translation process. The study of specific cases of indirect translation and further systematisation of the identified distortions allows us to track negative trends, and the novel Max Havelaar and its translations into different languages can serve as material for identifying these patterns.

When translating specific words from one language to another, it is necessary to take into consideration several different variables to guarantee correctness, precision, quality, and contextual appropriateness. Context-aware translation systems face several issues due to the complexities of natural language. These challenges include contextual disambiguation, cultural and linguistic nuances, subject matter expertise, and maintaining translation consistency. In low-resource languages such as Tamil, which suffer from limited linguistic resources, high word-sense ambiguity, and flexible word order, the translation process is further complicated. These issues are magnified in low-resource languages. To address these concerns, the proposed study will provide linguistic resources for pre-training Neural Machine Translation (NMT) systems. This will be accomplished by developing a translation system that is language-independent, document-specific, and context-based, making it suitable for languages with limited resources. The research presents a novel word vector format known as Word2Line (W2L), which minimises the time required for the process. With the use of TF-IDF, the system can recognise document-specific words in the English language, which serves as the source language, and then predict the context-based translations of those words in Tamil, the target language.

The Old Man and the Sea is a literature work by Ernest Hemingway. It has been translated into many languages even in Indonesian by Deera Army. Hemingway used more clause complex in producing his works. It causes problems in translation such as: translators should give more attention to the translation techniques used, readability decrease, and etc. On the other hand, Deera Army solved those problems by splitting the clause complex into shorter one. It is needed to conduct a study in how to make translation of complex clause. This study can be clearly conducted by using Systemic Functional Grammar (SFG) approach. In addition, this study is aimed at: (1) describing how can be interdependency and logical semantics of complex clause in source language realized into interdependency and logical semantics of complex clause in target language of The Old Man and the Sea Novel (2) describing what translation techniques on taxis markers are used in translating from source language to target language (3) describing translation quality of clause complex translation in target language. The result of the analysis showed that there are 400 sentences which have been broken into 701 clauses. Based on the analysis, there are paratactic and hypotactic form. Paratactic took 65.30% and hypotactic, 34.50%. All of them affect translation quality. Based on the analysis, the average of accuration takes up 2.89, naturalness with 2.96 and readibility with 2.97. The writer suggests that the next researcher can conduct the same research in the deeper way.

Pivoting through a popular language with more parallel corpora available (e.g. English and Chinese) is a common approach to build machine translation (MT) systems for low-resource languages. For example, to build a Russian-to-Spanish MT system, we could build one system using the Russian–Spanish corpus directly. We could also build two systems, Russian-to-English and English-to-Spanish, as the resources of the two language pairs are much larger than the Russian–Spanish pair, and use them cascadingly to translate texts in Russian into Spanish by pivoting through English. There are, however, some confusing results on the Pivot MT approach in the literature. In this paper, we reviewed the performance of Pivot MT with the United Nations Parallel Corpus v1.0 (UN6Way) using both English and Chinese as pivot languages. We also report our system performance on the CWMT 2018 Pivot MT shared task, where Japanese patent sentences are translated into English using Chinese as the pivot language.

Automatically assessing the quality of machine translation is an important yet challenging task for machine translation research. Translation quality assessment is understood as predicting translation quality without reference to the source text. Translation quality depends on the specific machine translation system and often requires post-editing. Manual editing is a long and expensive process. Since the need to quickly determine the quality of translation increases, its automation is required. In this paper, we propose a quality assessment method based on ensemble supervised machine learning methods. The bilingual corpus WMT 2019 for the EnglishRussian language pair was used as data. The text data volume is 17089 sentences, 85% of the data was used for training, and 15% for testing the model. Linguistic functions extracted from the text in the source and target languages were used as features for training the system, since it is these characteristics that can most accurately characterize the translation in terms of quality. The following tools were used for feature extraction: a free language modeling tool based on SRILM and a Stanford POS Tagger parts of speech tagger. Before training the system, the text was preprocessed. The model was trained using three regression methods: Bagging, Extra Tree, and Random Forest. The algorithms were implemented in the Python programming language using the Scikit learn library. The parameters of the random forest method have been optimized using a grid search. The performance of the model was assessed by the mean absolute error MAE and the root mean square error RMSE, as well as by the Pearsоn coefficient, which determines the correlation with human judgment. Testing was carried out using three machine translation systems: Google and Bing neural systems, Mouses statistical machine translation systems based on phrases and based on syntax. Based on the results of the work, the method of additional trees showed itself best. In addition, for all categories of indicators under consideration, the best results are achieved using the Google machine translation system. The developed method showed good results close to human judgment. The system can be used for further research in the task of assessing the quality of translation.

Most neural machine translation (NMT) systems rely on parallel data, comprising text in the source language and its corresponding translation in the target language. While it’s acknowledged that context enhances NMT models, this work proposes a novel approach by incorporating external context, specifically explanations of source text meanings, akin to how human translators leverage context for comprehension. The suggested methodology innovatively addresses the challenge of incorporating lengthy contextual information into NMT systems. By employing state-of-the-art transformer-based models, external context is integrated, thereby enriching the translation process. A key aspect of the approach lies in the utilization of diverse text summarization techniques, strategically employed to efficiently distill extensive contextual details into the NMT framework. This novel solution not only overcomes the obstacle posed by lengthy context but also enhances the translation quality, marking a advancement in the field of NMT. Furthermore, the data-centric approach ensures robustness and effectiveness, yielding improvements in translation quality, as evidenced by a considerable boost in BLEU score points ranging from 0.46 to 1.87 over baseline models. Additionally, we make our dataset publicly available, facilitating further research in this domain.

This paper investigates how multilingual language models benefit low-resource languages through our submission to the WMT 2025 Low-Resource Indic Language Translation shared task.We explore whether languages from related families can effectively support translation for low-resource languages that were absent or underrepresented during model training.Using a quantized multilingual pretrained foundation model, we examine zero-shot translation capabilities and cross-lingual transfer effects across three language families: Tibeto-Burman, Indo-Aryan, and Austroasiatic.Our findings demonstrate that multilingual models failed to leverage linguistic similarities, particularly evidenced within the Tibeto-Burman family.The study provides insights into the practical feasibility of zero-shot translation for low-resource language settings and the role of language family relationships in multilingual model performance.