Guard2Vul: Vulnerability Detection viaGradient-based Adversarial TrainingEnhanced Graph Learning

Guard2Vul: Vulnerability Detection viaGradient-based Adversarial TrainingEnhanced Graph Learning | 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 Guard2Vul: Vulnerability Detection viaGradient-based Adversarial TrainingEnhanced Graph Learning Hao Shen, Xiaolin Ju, Wenjie Li, Xiang Chen, Guang Yang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3840091/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 Modern software security is challenged by vulnerabilities, which canlead to severe consequences, including loss of information, property, andprivacy disclosure. Recently, graph-based deep learning methods havebeen proven as a promising solution for vulnerability detection at func-tional granularity. However, previous studies still face challenges, suchas structural imbalances, complex connections in abstract syntax treesof functions, and insufficient training data. In this study, we proposeGuard2Vul, which combines a graph neural network with a residualnetwork to jointly capture deeper semantic and structural features,specifically, source code and node dependencies. Moreover, we utilizeGraphSMOTE and DGD, a dropout-enhanced gradient adversarial train-ing technique, to conduct data augmentation and automatically improvenormalized stability. To demonstrate the effectiveness of Guard2Vul, we evaluate its performance on four experimental subjects by different pro-gramming languages, such as C/C++ and Java. To show the competitive-ness of Guard2Vul, we consider five Deep Learning-based vulnerabilitydetection approaches (i.e., TokenCNN, Sysevr, VulDeePecker, Devign,and Reveal) as baselines. The results indicate that Guard2Vul out-performs these baselines, achieving at least 13.4%, 28.9%, 6.2%, and12.1% higher F1-measure on four experimental subjects. Finally, weperform ablation experiments to demonstrate the effectiveness of ourcustomized components, namely enhanced graph representation learningand the gradient-based adversarial training method, in Guard2Vul. Vulnerability detection Graph convolutional neural network Abstract syntax tree Gradient-based adversarial training 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-3840091","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":266033634,"identity":"f56d542b-7795-4ae1-9397-b8d1b34da761","order_by":0,"name":"Hao Shen","email":"","orcid":"","institution":"Nantong University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Shen","suffix":""},{"id":266033635,"identity":"be7a7e62-51c1-4467-8af0-a38939f695ed","order_by":1,"name":"Xiaolin Ju","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAr0lEQVRIiWNgGAWjYBAC9gYg8aACwpEgSgsjSEvCGYhqErQktpGkZUYCm0TivDt1BgeYD97mYbDLI0YLs0HitmcSBgfYkq15GJKLidHC+CBx22GgFh4zaR6GA4kNRGgBKpsD0sL/jTgtgmBbGsC2sBGnRZrnYbNBwrHDkjMPsxlbzjFIJqyFjz35mMSHmsP8fMebH954U2FHWAs0ZoCAGUQYEFY/CkbBKBgFo4AIAABlkjfWJV3voQAAAABJRU5ErkJggg==","orcid":"","institution":"Nantong University","correspondingAuthor":true,"prefix":"","firstName":"Xiaolin","middleName":"","lastName":"Ju","suffix":""},{"id":266033636,"identity":"5d62c99b-94ce-4502-935a-5bf3e33c09bf","order_by":2,"name":"Wenjie Li","email":"","orcid":"","institution":"Anhui Normal University","correspondingAuthor":false,"prefix":"","firstName":"Wenjie","middleName":"","lastName":"Li","suffix":""},{"id":266033637,"identity":"bd1b21c2-0369-4acf-8303-e3b5e9cc5218","order_by":3,"name":"Xiang Chen","email":"","orcid":"","institution":"Nantong University","correspondingAuthor":false,"prefix":"","firstName":"Xiang","middleName":"","lastName":"Chen","suffix":""},{"id":266033638,"identity":"55296cae-283b-44ac-aa73-7b46e5655e60","order_by":4,"name":"Guang Yang","email":"","orcid":"","institution":"Nanjing University of Aeronautics and Astronautics","correspondingAuthor":false,"prefix":"","firstName":"Guang","middleName":"","lastName":"Yang","suffix":""}],"badges":[],"createdAt":"2024-01-06 15:29:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3840091/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3840091/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":64032365,"identity":"73c8796b-9cd2-4384-8498-5f53bc265e8c","added_by":"auto","created_at":"2024-09-05 09:33:45","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":582958,"visible":true,"origin":"","legend":"","description":"","filename":"ASEGuard2Vul.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3840091/v1_covered_597d1b50-703b-47d7-9874-99819d69556f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Guard2Vul: Vulnerability Detection viaGradient-based Adversarial TrainingEnhanced Graph Learning","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":"Vulnerability detection, Graph convolutional neural network, Abstract syntax tree, Gradient-based adversarial training","lastPublishedDoi":"10.21203/rs.3.rs-3840091/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3840091/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Modern software security is challenged by vulnerabilities, which canlead to severe consequences, including loss of information, property, andprivacy disclosure. Recently, graph-based deep learning methods havebeen proven as a promising solution for vulnerability detection at func-tional granularity. However, previous studies still face challenges, suchas structural imbalances, complex connections in abstract syntax treesof functions, and insufficient training data. In this study, we proposeGuard2Vul, which combines a graph neural network with a residualnetwork to jointly capture deeper semantic and structural features,specifically, source code and node dependencies. Moreover, we utilizeGraphSMOTE and DGD, a dropout-enhanced gradient adversarial train-ing technique, to conduct data augmentation and automatically improvenormalized stability. To demonstrate the effectiveness of Guard2Vul, we evaluate its performance on four experimental subjects by different pro-gramming languages, such as C/C++ and Java. To show the competitive-ness of Guard2Vul, we consider five Deep Learning-based vulnerabilitydetection approaches (i.e., TokenCNN, Sysevr, VulDeePecker, Devign,and Reveal) as baselines. The results indicate that Guard2Vul out-performs these baselines, achieving at least 13.4%, 28.9%, 6.2%, and12.1% higher F1-measure on four experimental subjects. Finally, weperform ablation experiments to demonstrate the effectiveness of ourcustomized components, namely enhanced graph representation learningand the gradient-based adversarial training method, in Guard2Vul.","manuscriptTitle":"Guard2Vul: Vulnerability Detection viaGradient-based Adversarial TrainingEnhanced Graph Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-10 05:21:47","doi":"10.21203/rs.3.rs-3840091/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":"64f4a9e8-9329-4b87-9ede-1947344b6873","owner":[],"postedDate":"January 10th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-09-05T09:25:38+00:00","versionOfRecord":[],"versionCreatedAt":"2024-01-10 05:21:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3840091","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3840091","identity":"rs-3840091","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}