An Optimized Approach for Handling Limited Vocabulary Constraints and Word Order Divergence in English-Urdu Neural Machine Translation

An Optimized Approach for Handling Limited Vocabulary Constraints and Word Order Divergence in English-Urdu Neural Machine Translation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article An Optimized Approach for Handling Limited Vocabulary Constraints and Word Order Divergence in English-Urdu Neural Machine Translation Rabail Asghar, Yella Mehroze, Zaheer Ahmad Gondal, Junaid Akram This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9389246/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Machine Translation (MT) in Natural Language Processing (NLP) seeks to bridge language barriers by developing automated systems capable of translating human languages. Transformer-based Neural Machine Translation (NMT) has surpassed traditional statistical and rule-based approaches in capturing complex linguistic structures and producing high-quality translations. Nevertheless, NMT still faces challenges with low-resource, morphologically rich language pairs such as English–Urdu, particularly in handling out-of-vocabulary (OOV) words and syntactic divergences. This study presents a structured pipeline to address these challenges. Publicly available English and Urdu corpora were collected and processed to train subword tokenizers, including Byte-Pair Encoding (BPE) and Sen-tencePiece variants. An English–Urdu translation dataset was prepared, cleaned, and made publicly available on Mendeley Data for reproducibility. Each tokenized dataset was used to train independent Transformer models, and the best-performing tokenizer–model combination was further fine-tuned by optimizing the positional encoding scaling parameter (λ). Experimental results demonstrate that this approach achieves substantial improvements over existing state-of-the-art English–Urdu MT systems , as reflected in BLEU scores. The findings highlight the effectiveness of combining subword tokenization with optimized positional encoding, providing a strong foundation for future research in low-resource neural machine translation. Neural Machine Translation (NMT) Natural Language Processing (NLP) English-Urdu Machine Translation (MT) Transformer Out-of-Vocabulary (OOV) words Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9389246","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627459271,"identity":"2fc1d5c1-6341-41c8-88db-50868da6a131","order_by":0,"name":"Rabail Asghar","email":"","orcid":"","institution":"COMSATS University Islamabad","correspondingAuthor":false,"prefix":"","firstName":"Rabail","middleName":"","lastName":"Asghar","suffix":""},{"id":627459272,"identity":"02ec7ddc-5edd-41af-8ac1-675abfbd5616","order_by":1,"name":"Yella Mehroze","email":"","orcid":"","institution":"COMSATS University Islamabad","correspondingAuthor":false,"prefix":"","firstName":"Yella","middleName":"","lastName":"Mehroze","suffix":""},{"id":627459273,"identity":"511af26c-0bf6-417f-8801-3a7d036b2272","order_by":2,"name":"Zaheer Ahmad Gondal","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Zaheer","middleName":"Ahmad","lastName":"Gondal","suffix":""},{"id":627459274,"identity":"50605178-e447-42ed-ae1a-33470f7039cc","order_by":3,"name":"Junaid Akram","email":"data:image/png;base64,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","orcid":"","institution":"COMSATS University Islamabad","correspondingAuthor":true,"prefix":"","firstName":"Junaid","middleName":"","lastName":"Akram","suffix":""}],"badges":[],"createdAt":"2026-04-11 15:53:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9389246/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9389246/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107870518,"identity":"99a7f270-f94a-4dc6-8d14-51f9744fc370","added_by":"auto","created_at":"2026-04-27 07:39:50","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":346705,"visible":true,"origin":"","legend":"","description":"","filename":"VocabularyConstraintsHandling.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9389246/v1_covered_138c7018-f1e5-4771-a55b-ff8cdd3f7b86.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"An Optimized Approach for Handling Limited Vocabulary Constraints and Word Order Divergence in English-Urdu Neural Machine Translation","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Neural Machine Translation (NMT), Natural Language Processing (NLP), English-Urdu Machine Translation (MT), Transformer, Out-of-Vocabulary (OOV) words","lastPublishedDoi":"10.21203/rs.3.rs-9389246/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9389246/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Machine Translation (MT) in Natural Language Processing (NLP) seeks to bridge language barriers by developing automated systems capable of translating human languages. Transformer-based Neural Machine Translation (NMT) has surpassed traditional statistical and rule-based approaches in capturing complex linguistic structures and producing high-quality translations. Nevertheless, NMT still faces challenges with low-resource, morphologically rich language pairs such as English–Urdu, particularly in handling out-of-vocabulary (OOV) words and syntactic divergences. This study presents a structured pipeline to address these challenges. Publicly available English and Urdu corpora were collected and processed to train subword tokenizers, including Byte-Pair Encoding (BPE) and Sen-tencePiece variants. An English–Urdu translation dataset was prepared, cleaned, and made publicly available on Mendeley Data for reproducibility. Each tokenized dataset was used to train independent Transformer models, and the best-performing tokenizer–model combination was further fine-tuned by optimizing the positional encoding scaling parameter (λ). Experimental results demonstrate that this approach achieves substantial improvements over existing state-of-the-art English–Urdu MT systems , as reflected in BLEU scores. The findings highlight the effectiveness of combining subword tokenization with optimized positional encoding, providing a strong foundation for future research in low-resource neural machine translation.","manuscriptTitle":"An Optimized Approach for Handling Limited Vocabulary Constraints and Word Order Divergence in English-Urdu Neural Machine Translation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-27 04:06:59","doi":"10.21203/rs.3.rs-9389246/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"aec8751a-ad6d-432a-8338-db17f70b7e40","owner":[],"postedDate":"April 27th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-17T07:28:12+00:00","index":17,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-02T22:48:03+00:00","index":16,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-27T04:06:59+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-27 04:06:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9389246","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9389246","identity":"rs-9389246","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}