Learning Disentangled Multi-intent Representations for Scalable Recommendation

Learning Disentangled Multi-intent Representations for Scalable Recommendation | 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 Learning Disentangled Multi-intent Representations for Scalable Recommendation Zhandong Mei, Yongyi Lin, Shanfan Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6259280/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 29 Apr, 2026 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted 11 You are reading this latest preprint version Abstract Intent learning aims to capture user intentions to enhance user understanding and item recommendation, emerging as a key research focus in recent years. However, existing approaches often rely on complex and computationally expensive alternating optimization processes, which limit both performance and scalability. To address these challenges, we propose a novel intent learning method, termed DMI , which unifies behavior representation learning within an end-to-end trainable clustering framework to achieve effective and efficient recommendations. Specifically, user behavior sequences are encoded, and cluster centers—representing latent intents—are initialized as learnable neural parameters. A novel learnable clustering module is designed to separate different cluster centers, thereby disentangling users' complex intentions. Moreover, by explicitly modeling the relationships between user and item intents, DMI enforces behavior embeddings to align with cluster centers in an end-to-end manner, guiding the network to extract user intent from behavioral data. Since DMI eliminates the need for full-graph training, it enables the simultaneous optimization of recommendation and clustering through mini-batch updates. By integrating both global and local learning, DMI effectively captures high-quality node embeddings. Compared to the runner-up model, DMI significantly reduces computational costs while improving NDCG@3 by 13.68% on the DBLP dataset. Recommender systems Multiple intents Disentangled Representation Clustering Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 29 Apr, 2026 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted Editorial decision: Revision requested 10 Apr, 2025 Reviews received at journal 10 Apr, 2025 Reviews received at journal 28 Mar, 2025 Reviewers agreed at journal 28 Mar, 2025 Reviews received at journal 24 Mar, 2025 Reviewers agreed at journal 21 Mar, 2025 Reviewers agreed at journal 21 Mar, 2025 Reviewers invited by journal 21 Mar, 2025 Editor assigned by journal 20 Mar, 2025 Submission checks completed at journal 20 Mar, 2025 First submitted to journal 19 Mar, 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-6259280","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":435357070,"identity":"0ae0052f-f39c-45fe-94ba-c842b947de5f","order_by":0,"name":"Zhandong Mei","email":"","orcid":"","institution":"Xi’an Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Zhandong","middleName":"","lastName":"Mei","suffix":""},{"id":435357071,"identity":"c37a2f6d-d297-49e7-8817-ca9f3dc068e8","order_by":1,"name":"Yongyi Lin","email":"","orcid":"","institution":"Xi’an Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Yongyi","middleName":"","lastName":"Lin","suffix":""},{"id":435357072,"identity":"0827e127-8307-4ba6-8d51-5cfe7d89a3e1","order_by":2,"name":"Shanfan Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/klEQVRIiWNgGAWjYDACCTApB8QHGBg+ACk2duK0GAPxYQbGGSAtzMRrYWZg5mEA0/iB/OzmZw+//DFI7G88f0za5tc2eT5mBsYPH3Nwa2Gcc8zcWLbNwFjiwGE26dy+24ZtQNskZ27DrYVZIsFMWrLhjxwDWEvPbUagFjZmXjxa2CTSv0lL/DHgkQdpsey5bU9QC49EjpnkBzYDOQOQFoYftxMJapGQyCmTZgT6xfDAYWPL3obbyW3MjM14/SI/I32b5A9giM27cfDhjR9/btvOb28++OEjHi3gIABHh8QBYPi1gViMDfjVg5T8AJH8IIV/CCoeBaNgFIyCEQgA0IpNwL8IRnQAAAAASUVORK5CYII=","orcid":"","institution":"Xi’an Jiaotong University","correspondingAuthor":true,"prefix":"","firstName":"Shanfan","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-03-19 08:08:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6259280/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6259280/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s13042-026-03125-1","type":"published","date":"2026-04-29T15:58:03+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":108437803,"identity":"6569591f-e841-4bd2-afb2-d76790e0ae08","added_by":"auto","created_at":"2026-05-04 16:03:38","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1343889,"visible":true,"origin":"","legend":"","description":"","filename":"SpringerNatureLaTeXTemplate.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6259280/v1_covered_8281188c-2166-4c84-bdc5-7677142bfb42.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Learning Disentangled Multi-intent Representations for Scalable Recommendation","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Recommender systems, Multiple intents, Disentangled Representation, Clustering","lastPublishedDoi":"10.21203/rs.3.rs-6259280/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6259280/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIntent learning aims to capture user intentions to enhance user understanding and item recommendation, emerging as a key research focus in recent years. However, existing approaches often rely on complex and computationally expensive alternating optimization processes, which limit both performance and scalability. To address these challenges, we propose a novel intent learning method, termed \u003cem\u003eDMI\u003c/em\u003e, which unifies behavior representation learning within an end-to-end trainable clustering framework to achieve effective and efficient recommendations. Specifically, user behavior sequences are encoded, and cluster centers—representing latent intents—are initialized as learnable neural parameters. A novel learnable clustering module is designed to separate different cluster centers, thereby disentangling users' complex intentions. Moreover, by explicitly modeling the relationships between user and item intents, \u003cem\u003eDMI\u003c/em\u003e enforces behavior embeddings to align with cluster centers in an end-to-end manner, guiding the network to extract user intent from behavioral data. Since \u003cem\u003eDMI\u003c/em\u003e eliminates the need for full-graph training, it enables the simultaneous optimization of recommendation and clustering through mini-batch updates. By integrating both global and local learning, \u003cem\u003eDMI\u003c/em\u003e effectively captures high-quality node embeddings. Compared to the runner-up model, \u003cem\u003eDMI\u003c/em\u003e significantly reduces computational costs while improving NDCG@3 by 13.68% on the \u003cem\u003eDBLP\u003c/em\u003e dataset.\u003c/p\u003e","manuscriptTitle":"Learning Disentangled Multi-intent Representations for Scalable Recommendation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-31 04:02:53","doi":"10.21203/rs.3.rs-6259280/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-04-11T03:48:55+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-10T10:20:26+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-28T12:46:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"221430011060436093875427485572631615892","date":"2025-03-28T10:18:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-24T14:22:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"24317964155321407631571747260412620232","date":"2025-03-22T02:58:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"120306718412046508958782721044923374642","date":"2025-03-21T14:47:20+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-21T13:56:43+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-21T01:15:33+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-20T11:53:14+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Machine Learning and Cybernetics","date":"2025-03-19T08:05:54+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"849d6d19-07a4-43d9-b387-4752dc759f6b","owner":[],"postedDate":"March 31st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-05-04T16:02:47+00:00","versionOfRecord":{"articleIdentity":"rs-6259280","link":"https://doi.org/10.1007/s13042-026-03125-1","journal":{"identity":"international-journal-of-machine-learning-and-cybernetics","isVorOnly":false,"title":"International Journal of Machine Learning and Cybernetics"},"publishedOn":"2026-04-29 15:58:03","publishedOnDateReadable":"April 29th, 2026"},"versionCreatedAt":"2025-03-31 04:02:53","video":"","vorDoi":"10.1007/s13042-026-03125-1","vorDoiUrl":"https://doi.org/10.1007/s13042-026-03125-1","workflowStages":[]},"version":"v1","identity":"rs-6259280","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6259280","identity":"rs-6259280","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}