Dynamic Graph Anomaly Detection via Temporal-Structural Attention and Variational Graph Autoencoder

Dynamic Graph Anomaly Detection via Temporal-Structural Attention and Variational Graph Autoencoder | 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 Dynamic Graph Anomaly Detection via Temporal-Structural Attention and Variational Graph Autoencoder Jinyong Sun, Zhiliang Shang, Zhigang Sun, Xiang Zhao, Guoyong Cai, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7806501/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract With the increasing application of dynamic graph data in social networks, financial services and cybersecurity, dynamic graph anomaly detection has attracted growing attention. Due to the dynamicity and complexity arising from the continuous evolution of dynamic graph topology, existing methods face challenges in effectively modeling complex temporal features and capturing the intricate dependencies between temporal and structural information. These challenges hinder the comprehensive capture of spatiotemporal features in dynamic graphs, thereby limiting anomaly detection performance. To address these issues, we propose a novel dynamic graph anomaly detection method that integrates Temporal-Structural Attention graph embedding with Variational Graph Autoencoder (TSAVGA). Our approach innovatively employs a hierarchical temporal-structural attention mechanism to capture short-term spatiotemporal features, while leveraging a GCN-GRU enhanced variational graph autoencoder to model deeper long-term dependencies. In addition, we design a spectral clustering-based anomaly injection strategy to augment the diversity of anomalies in the training data. Experimental results on six real-world datasets demonstrate that TSAVGA outperforms state-of-the-art methods by 1% to 5% in AUC, and maintains robust performance across varying anomaly ratios, confirming both the effectiveness and stability of our method. Dynamic graphs Anomaly detection Attention mechanism Variational graph autoencoder Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 01 May, 2026 Reviews received at journal 27 Jan, 2026 Reviews received at journal 22 Jan, 2026 Reviews received at journal 22 Jan, 2026 Reviewers agreed at journal 12 Jan, 2026 Reviewers agreed at journal 07 Jan, 2026 Reviewers agreed at journal 05 Jan, 2026 Reviewers invited by journal 05 Jan, 2026 Editor assigned by journal 04 Nov, 2025 Submission checks completed at journal 15 Oct, 2025 First submitted to journal 08 Oct, 2025 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-7806501","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":569289789,"identity":"06cfd4c2-f8f8-4d82-bb4e-2ce6e793c8b9","order_by":0,"name":"Jinyong Sun","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Jinyong","middleName":"","lastName":"Sun","suffix":""},{"id":569289791,"identity":"60f94488-50a9-46ac-985f-b6ad2d10abc1","order_by":1,"name":"Zhiliang Shang","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhiliang","middleName":"","lastName":"Shang","suffix":""},{"id":569289793,"identity":"8de76785-f82b-4964-960b-07d2257264c1","order_by":2,"name":"Zhigang Sun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxUlEQVRIiWNgGAWjYNCCCijNQ7yWMyRrYWwjRYt8/xmzh1/n1SZul0hgfPC2jUHenJAWgwNnzI1ltx03tpyRwGw4t43BcGcDIS2MPWbSktuOyRncSGCT5m1jSDA4QMhhzTxALXOO8QC1sP8mSgvDMR4zyY8NNWBbmInSYnCGrUya4dgBY4MzD5sl55yTMNxA0GH9h7dJ/qipS9xwPPnghzdlNvKEHQYEzDwMh4EUYwOQkCBCPUjtD4Y64lSOglEwCkbByAQAaUo8xSbSX0sAAAAASUVORK5CYII=","orcid":"","institution":"Guangxi Normal University","correspondingAuthor":true,"prefix":"","firstName":"Zhigang","middleName":"","lastName":"Sun","suffix":""},{"id":569289795,"identity":"71d4c71f-f4ad-4a85-a9eb-bade3e8ceab9","order_by":3,"name":"Xiang Zhao","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiang","middleName":"","lastName":"Zhao","suffix":""},{"id":569289798,"identity":"9b349c31-d957-43d6-8577-8aadac6c2047","order_by":4,"name":"Guoyong Cai","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Guoyong","middleName":"","lastName":"Cai","suffix":""},{"id":569289803,"identity":"8daf66f1-9fd3-4bd9-add2-746b52c5720d","order_by":5,"name":"Ruiheng Ling","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Ruiheng","middleName":"","lastName":"Ling","suffix":""},{"id":569289805,"identity":"269bbf87-0762-4254-b641-cfd7bbf88cc6","order_by":6,"name":"Chuang Lei","email":"","orcid":"","institution":"Guilin University of Electronic Technology","correspondingAuthor":false,"prefix":"","firstName":"Chuang","middleName":"","lastName":"Lei","suffix":""}],"badges":[],"createdAt":"2025-10-08 09:53:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7806501/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7806501/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99793767,"identity":"b7985de7-21fa-42fc-8e84-8967a9788a68","added_by":"auto","created_at":"2026-01-08 13:32:20","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8760,"visible":true,"origin":"","legend":"","description":"","filename":"fe54a319accd4de79bcd13b63be04d74.json","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/04be7c652addff0b453cc867.json"},{"id":99615647,"identity":"13bff76a-90b2-4db1-aa2c-fa77826a7065","added_by":"auto","created_at":"2026-01-06 13:16:21","extension":"xml","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":113312,"visible":true,"origin":"","legend":"","description":"","filename":"fe54a319accd4de79bcd13b63be04d741enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/3fb7c34f0df5f2210343a7db.xml"},{"id":99615648,"identity":"c70a6057-4b04-4321-88db-16c3b80ef41d","added_by":"auto","created_at":"2026-01-06 13:16:21","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71417,"visible":true,"origin":"","legend":"","description":"","filename":"CoverLetter.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/c4e6631d271802aaa725159a.pdf"},{"id":99615651,"identity":"3e7a0435-f8e2-430a-82ed-54c7b508f03f","added_by":"auto","created_at":"2026-01-06 13:16:21","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1729952,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/6dc83a0325e863b95319a24b.pdf"},{"id":99615653,"identity":"a0475c6c-18ae-40aa-b0ce-30fa22657145","added_by":"auto","created_at":"2026-01-06 13:16:21","extension":"cls","order_by":12,"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-7806501/v1/cb94b5f13ce0ede3d358e103.cls"},{"id":99793486,"identity":"72ad6e68-2881-401d-82e3-02a42bca95b3","added_by":"auto","created_at":"2026-01-08 13:31:40","extension":"bst","order_by":13,"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-7806501/v1/88b948f3d2306dffc6dd7385.bst"},{"id":99794018,"identity":"31a4f789-6d40-464e-8087-6a3372ac5b11","added_by":"auto","created_at":"2026-01-08 13:33:47","extension":"xml","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":118712,"visible":true,"origin":"","legend":"","description":"","filename":"fe54a319accd4de79bcd13b63be04d741structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/d569b6a37a2412e0adc2be0f.xml"},{"id":99615654,"identity":"fa7358bf-cd59-4d64-99c4-00616cd79557","added_by":"auto","created_at":"2026-01-06 13:16:22","extension":"html","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137658,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1/52b82a38a89158b06365c0c0.html"},{"id":99804311,"identity":"bc4b44f2-de12-4079-a369-1cbd57a646e7","added_by":"auto","created_at":"2026-01-08 14:13:03","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1631548,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7806501/v1_covered_38db4540-870c-4980-ba1d-61b722bb9f40.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Dynamic Graph Anomaly Detection via Temporal-Structural Attention and Variational Graph Autoencoder","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"cluster-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Cluster Computing](https://www.springer.com/journal/10586)","snPcode":"10586","submissionUrl":"https://submission.nature.com/new-submission/10586/3","title":"Cluster Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Dynamic graphs, Anomaly detection, Attention mechanism, Variational graph autoencoder","lastPublishedDoi":"10.21203/rs.3.rs-7806501/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7806501/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWith the increasing application of dynamic graph data in social networks, financial services and cybersecurity, dynamic graph anomaly detection has attracted growing attention. Due to the dynamicity and complexity arising from the continuous evolution of dynamic graph topology, existing methods face challenges in effectively modeling complex temporal features and capturing the intricate dependencies between temporal and structural information. These challenges hinder the comprehensive capture of spatiotemporal features in dynamic graphs, thereby limiting anomaly detection performance. To address these issues, we propose a novel dynamic graph anomaly detection method that integrates Temporal-Structural Attention graph embedding with Variational Graph Autoencoder (TSAVGA). Our approach innovatively employs a hierarchical temporal-structural attention mechanism to capture short-term spatiotemporal features, while leveraging a GCN-GRU enhanced variational graph autoencoder to model deeper long-term dependencies. In addition, we design a spectral clustering-based anomaly injection strategy to augment the diversity of anomalies in the training data. Experimental results on six real-world datasets demonstrate that TSAVGA outperforms state-of-the-art methods by 1% to 5% in AUC, and maintains robust performance across varying anomaly ratios, confirming both the effectiveness and stability of our method.\u003c/p\u003e","manuscriptTitle":"Dynamic Graph Anomaly Detection via Temporal-Structural Attention and Variational Graph Autoencoder","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-06 13:16:16","doi":"10.21203/rs.3.rs-7806501/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-01T11:08:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-27T08:09:34+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-22T12:48:08+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-22T07:12:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"81018855626956109820446472115451398281","date":"2026-01-13T01:41:11+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"178075945418771984241951476066326397996","date":"2026-01-07T22:47:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"24827273141904162918269132555199900526","date":"2026-01-05T06:47:26+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-05T06:27:17+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-04T06:30:06+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-15T11:50:58+00:00","index":"","fulltext":""},{"type":"submitted","content":"Cluster Computing","date":"2025-10-08T09:40:17+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"cluster-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Cluster Computing](https://www.springer.com/journal/10586)","snPcode":"10586","submissionUrl":"https://submission.nature.com/new-submission/10586/3","title":"Cluster Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"65b9cdf0-239a-492e-b326-1df9ff2814ad","owner":[],"postedDate":"January 6th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-01T11:08:54+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-01T11:23:57+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-06 13:16:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7806501","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7806501","identity":"rs-7806501","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}