Few-Shot Supervised OCR Verification with Graph Convolutional Networks on Relational Line Graphs

Few-Shot Supervised OCR Verification with Graph Convolutional Networks on Relational Line Graphs | 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 Few-Shot Supervised OCR Verification with Graph Convolutional Networks on Relational Line Graphs Shikhar Dubey, Manikandan Ravikiran, Rohit Saluja This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8121737/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 Optical Character Recognition (OCR) systems are foundational to document digitization, yet their outputs often contain errors that require costly manual verification. Most automated verification methods are supervised, demanding large, meticulously labeled datasets of correct and incorrect text, which presents a significant bottleneck. In this paper, we propose a novel few-shot supervised framework for post-hoc OCR verification using Graph Neural Networks (GNNs). Our primary contribution is a unique graph-based representation of a word. We model each word as a line graph, where each node corresponds to a relational comparison—either between a real character and its synthetic counterpart ( \((\text{R}_i)\) - \((\text{S}_i)\) ) or between two sequential real characters ( \((\text{R}_i)\) - \((\text{R}_{i+1})\) ). These nodes are encoded with a rich feature vector combining classic geometric properties (Hu Moments, pixel counts) with deep visual features and a cross-entropy stability metric derived from data augmentation. We frame verification as a few-shot graph classification task, benchmarking GCN, GAT, and GraphSAGE architectures across varying depths on a dataset collected from approximately 50 archival books. Contrary to the assumption that deeper models yield better representations, our results demonstrate that a lightweight 3-layer Graph Convolutional Network (GCN) outperforms deeper 5-layer variants and attention-based models. With as few as 50 labeled examples, this compact model achieves competitive error detection performance, establishing a highly effective and scalable alternative for OCR verification that mitigates the dependence on large-scale labeled data. Optical Character Recognition (OCR) Graph Convolutional Networks (GCNs) Word Graph Representation OCR Verification 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-8121737","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":550751753,"identity":"030a3056-6089-4991-a6cc-7747035d2091","order_by":0,"name":"Shikhar Dubey","email":"data:image/png;base64,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","orcid":"","institution":"Indian Institute of Technology Mandi","correspondingAuthor":true,"prefix":"","firstName":"Shikhar","middleName":"","lastName":"Dubey","suffix":""},{"id":550751754,"identity":"17d73b21-8653-4330-808b-8085bb6b76fa","order_by":1,"name":"Manikandan Ravikiran","email":"","orcid":"","institution":"Indian Institute of Technology Mandi","correspondingAuthor":false,"prefix":"","firstName":"Manikandan","middleName":"","lastName":"Ravikiran","suffix":""},{"id":550751755,"identity":"f6afed4d-8809-428a-9ffb-12706933c813","order_by":2,"name":"Rohit Saluja","email":"","orcid":"","institution":"Indian Institute of Technology Mandi","correspondingAuthor":false,"prefix":"","firstName":"Rohit","middleName":"","lastName":"Saluja","suffix":""}],"badges":[],"createdAt":"2025-11-15 11:38:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8121737/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8121737/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97109327,"identity":"1488e6ed-59e9-4a52-be8c-f31603d07276","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5112,"visible":true,"origin":"","legend":"","description":"","filename":"dcd14d1bcf224415be1c65ed9ff6bb6a.json","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/4f03c64d90f6ab7cb4987917.json"},{"id":97109330,"identity":"bbbef1a9-9cd4-4222-abf3-10ba8e062299","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"xml","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64402,"visible":true,"origin":"","legend":"","description":"","filename":"dcd14d1bcf224415be1c65ed9ff6bb6a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/04424fd9c531516f27b62668.xml"},{"id":97109338,"identity":"d8eecc3a-adfd-49b6-81ae-7140db0891da","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":329876,"visible":true,"origin":"","legend":"","description":"","filename":"Architecture.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/a9d61d3cc14daf8088ca2204.png"},{"id":97109335,"identity":"30724c88-b248-4dab-a68e-b306d732fd46","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":332004,"visible":true,"origin":"","legend":"","description":"","filename":"Architecture1.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/64d5f9eed78945bf2172b611.png"},{"id":97109343,"identity":"736967ee-4742-410e-a634-1051b7f23626","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69976,"visible":true,"origin":"","legend":"","description":"","filename":"COVERLETTER.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/282742a990075dd9445af93c.pdf"},{"id":97109349,"identity":"e0385bf5-9b9d-42d0-8ee8-886708949fe9","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":821833,"visible":true,"origin":"","legend":"","description":"","filename":"Datasetpages.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/60eefa90822ebfa5d6610a47.png"},{"id":97109353,"identity":"754cfc87-3d08-42ab-905f-e88c4baee4f7","added_by":"auto","created_at":"2025-12-01 05:54:05","extension":"pdf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1117819,"visible":true,"origin":"","legend":"","description":"","filename":"ShikharIJDAR2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/2e657c86a3b4fd267f8fd376.pdf"},{"id":97140221,"identity":"a1d86983-2ce2-4939-872e-e8f90c793af8","added_by":"auto","created_at":"2025-12-01 10:04:11","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":44508,"visible":true,"origin":"","legend":"","description":"","filename":"correctgood.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/a8460728b6767e55d196320f.png"},{"id":97109326,"identity":"8768ad15-226a-4472-bd1d-2ca914cbe76a","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"eps","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2890,"visible":true,"origin":"","legend":"","description":"","filename":"empty.eps","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/ee3e5af2a5f498ed43bd8d5d.eps"},{"id":97109341,"identity":"82026700-6ec3-44bd-abac-293c2ff91ca8","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"eps","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":91593,"visible":true,"origin":"","legend":"","description":"","filename":"fig.eps","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/8e2922913806bc9251ee5749.eps"},{"id":97142445,"identity":"4b4c3093-9b15-4a92-8600-494cc9b54683","added_by":"auto","created_at":"2025-12-01 10:07:37","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":67586,"visible":true,"origin":"","legend":"","description":"","filename":"incorrectgood.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/365de3f243d44ae78d050814.png"},{"id":97109339,"identity":"3a1a9734-99b5-49c8-b28d-079f0fb5e72e","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":66377,"visible":true,"origin":"","legend":"","description":"","filename":"incorrectwrong.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/1f097b77163f3e6397bb15e9.png"},{"id":97109337,"identity":"17b2c15a-6839-490b-b17b-f9c03e7e4b81","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"jpg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":945459,"visible":true,"origin":"","legend":"","description":"","filename":"linelevelsegmentationpipeline.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/d85a3cbcee19bf90cf226dd8.jpg"},{"id":97109350,"identity":"9e298d05-ffbe-4dc7-a53a-21b870345a0f","added_by":"auto","created_at":"2025-12-01 05:54:05","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54483,"visible":true,"origin":"","legend":"","description":"","filename":"placeholdergraphcorrect.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/528ad03c22776f59ff350860.png"},{"id":97109332,"identity":"16422de2-1d31-46ad-8153-877f0b613de9","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":72738,"visible":true,"origin":"","legend":"","description":"","filename":"placeholdergraphincorrect.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/8d8642515cd2cd8710e65aee.png"},{"id":97109328,"identity":"2c0d6263-d5a6-4a77-9ec6-d1ccfdedd0bb","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"bst","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":146013,"visible":true,"origin":"","legend":"","description":"","filename":"snapacite.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/443c63f28e45352335eebea6.bst"},{"id":97109346,"identity":"42adba95-8c7d-4986-8f1a-f8a3d242aaf1","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"bst","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29828,"visible":true,"origin":"","legend":"","description":"","filename":"snaps.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/481bad5304ee230e05a4532e.bst"},{"id":97109324,"identity":"ecf7ce06-8e43-44f4-aa01-c9e891a149a5","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"pdf","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":421391,"visible":true,"origin":"","legend":"","description":"","filename":"snarticle.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/0f6d8efba2bc64d6fe26788b.pdf"},{"id":97142214,"identity":"8f3fd117-c8ec-4ead-944e-5bf3ff3b2b3c","added_by":"auto","created_at":"2025-12-01 10:07:25","extension":"bst","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35515,"visible":true,"origin":"","legend":"","description":"","filename":"snbasic.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/b08e7e7f1543629c140b4d30.bst"},{"id":97109334,"identity":"0b78ee99-a769-4498-9423-d927623a6d2f","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"bst","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":33968,"visible":true,"origin":"","legend":"","description":"","filename":"snchicago.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/c0ed8b3f4cfb7a79b56627f0.bst"},{"id":97141599,"identity":"852a7031-c6cb-45e8-8537-7c6f634596d2","added_by":"auto","created_at":"2025-12-01 10:06:50","extension":"cls","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55857,"visible":true,"origin":"","legend":"","description":"","filename":"snjnl.cls","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/f2e0ba09695fdad9e9e9b7e5.cls"},{"id":97141195,"identity":"ec5a798f-ae6e-482e-98b3-c2a2c0cbf4de","added_by":"auto","created_at":"2025-12-01 10:06:24","extension":"bst","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64023,"visible":true,"origin":"","legend":"","description":"","filename":"snmathphysay.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/decaf32e1a2ed9a0a8b7d111.bst"},{"id":97109352,"identity":"54c8cfb0-dd53-4c6c-bfd6-970f744599cf","added_by":"auto","created_at":"2025-12-01 05:54:05","extension":"bst","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64166,"visible":true,"origin":"","legend":"","description":"","filename":"snmathphysnum.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/2d178b26a205187328f7e48a.bst"},{"id":97109325,"identity":"3f0bec0a-05ac-4816-ab0c-8b6807f62960","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"bst","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37333,"visible":true,"origin":"","legend":"","description":"","filename":"snnature.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/2e81044e4aa7c1e1e132dc99.bst"},{"id":97140485,"identity":"43617091-705a-4090-9aad-6c91a298f9a3","added_by":"auto","created_at":"2025-12-01 10:05:06","extension":"bst","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39951,"visible":true,"origin":"","legend":"","description":"","filename":"snvancouveray.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/75875326b479910730d1db1e.bst"},{"id":97109323,"identity":"023e81d1-35c8-49a5-9056-514bf15343b0","added_by":"auto","created_at":"2025-12-01 05:54:02","extension":"bst","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40758,"visible":true,"origin":"","legend":"","description":"","filename":"snvancouvernum.bst","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/5ac039e5aee683d9c923395b.bst"},{"id":97109333,"identity":"4a280923-a37d-4ac0-99c4-a8477fb6788e","added_by":"auto","created_at":"2025-12-01 05:54:03","extension":"pdf","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":418495,"visible":true,"origin":"","legend":"","description":"","filename":"usermanual.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/b71c0a3be3b2eab77495d73c.pdf"},{"id":97109354,"identity":"09c1c7cf-8d01-41b1-aaf7-b014b4e17883","added_by":"auto","created_at":"2025-12-01 05:54:05","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":243363,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineArchitecture1.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/6a7292dd65d6dca43fe19643.png"},{"id":97109348,"identity":"8f9fc6e3-9dd6-419d-8b8c-6e12d4442313","added_by":"auto","created_at":"2025-12-01 05:54:04","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3569415,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinelinelevelsegmentationpipeline.png","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/2e464e2af8fab29ebe361e1f.png"},{"id":97141020,"identity":"8d0268bc-475a-4f9d-9967-07e2bfa53d5a","added_by":"auto","created_at":"2025-12-01 10:06:08","extension":"xml","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70667,"visible":true,"origin":"","legend":"","description":"","filename":"dcd14d1bcf224415be1c65ed9ff6bb6a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/afa213df7ef183bafdc658fd.xml"},{"id":97141678,"identity":"c109688d-9c92-4519-962b-732333e73955","added_by":"auto","created_at":"2025-12-01 10:06:53","extension":"html","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":78991,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1/550a1ace3f1732dd83510998.html"},{"id":100905377,"identity":"f15c72f7-b57d-4499-871a-dbaa4efd40e2","added_by":"auto","created_at":"2026-01-22 15:41:54","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":459741,"visible":true,"origin":"","legend":"","description":"","filename":"ShikharIJDAR2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8121737/v1_covered_e4916cde-c9af-418a-8796-9b9f954d2b91.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Few-Shot Supervised OCR Verification with Graph Convolutional Networks on Relational Line Graphs","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":"Optical Character Recognition (OCR), Graph Convolutional Networks (GCNs), Word Graph Representation, OCR Verification","lastPublishedDoi":"10.21203/rs.3.rs-8121737/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8121737/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eOptical Character Recognition (OCR) systems are foundational to document digitization, yet their outputs often contain errors that require costly manual verification. Most automated verification methods are supervised, demanding large, meticulously labeled datasets of correct and incorrect text, which presents a significant bottleneck. In this paper, we propose a novel few-shot supervised framework for post-hoc OCR verification using Graph Neural Networks (GNNs). Our primary contribution is a unique graph-based representation of a word. We model each word as a line graph, where each node corresponds to a relational comparison\u0026mdash;either between a real character and its synthetic counterpart (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((\\text{R}_i)\\)\u003c/span\u003e\u003c/span\u003e-\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((\\text{S}_i)\\)\u003c/span\u003e\u003c/span\u003e) or between two sequential real characters (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((\\text{R}_i)\\)\u003c/span\u003e\u003c/span\u003e-\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((\\text{R}_{i+1})\\)\u003c/span\u003e\u003c/span\u003e). These nodes are encoded with a rich feature vector combining classic geometric properties (Hu Moments, pixel counts) with deep visual features and a cross-entropy stability metric derived from data augmentation. We frame verification as a few-shot graph classification task, benchmarking GCN, GAT, and GraphSAGE architectures across varying depths on a dataset collected from approximately 50 archival books. Contrary to the assumption that deeper models yield better representations, our results demonstrate that a lightweight 3-layer Graph Convolutional Network (GCN) outperforms deeper 5-layer variants and attention-based models. With as few as 50 labeled examples, this compact model achieves competitive error detection performance, establishing a highly effective and scalable alternative for OCR verification that mitigates the dependence on large-scale labeled data.\u003c/p\u003e","manuscriptTitle":"Few-Shot Supervised OCR Verification with Graph Convolutional Networks on Relational Line Graphs","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-01 05:53:52","doi":"10.21203/rs.3.rs-8121737/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":"707e811d-4f3f-43bf-a48d-d47db6546199","owner":[],"postedDate":"December 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-22T15:40:07+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-01 05:53:52","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8121737","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8121737","identity":"rs-8121737","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}