H-PLE: Algorithmic modeling of recommendation systems based on hierarchical PLE

H-PLE: Algorithmic modeling of recommendation systems based on hierarchical PLE | 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 H-PLE: Algorithmic modeling of recommendation systems based on hierarchical PLE Jinghao Xue, Tianxiang Yang, Hideo Suzuki This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8114131/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 In this study, we propose H-PLE (Hierarchical Progressive Layered Extraction), a novel multi-task recommendation framework designed to alleviate negative transfer and enhance performance under extreme abel imbalance. By introducing a hierarchical expert structure combined with improved gating and feature interaction modules, H-PLE effectively balances shared and task-specific knowledge extraction. We evaluate the model on large-scale real-world datasets and compare it against established baselines such as MMoE and PLE. The experimental analysis demonstrates that H-PLE consistently achieves stronger ranking performance, particularly in sparse-label tasks, while maintaining stable optimization behavior across multiple runs. Moreover, we reveal a practical gap between high ranking metrics and low decision-level metrics under default thresholding, and provide solutions through probability calibration, validation-based threshold selection, and top-K or rate-controlled strategies. These findings not only highlight the methodological advantages of H-PLE but also contribute actionable insights for bridging the gap between offline evaluation and online deployment in large-scale recommendation systems. Multitask Learning Recommendation Systems Hierarchical Modeling Negative Transfer 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-8114131","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":554856591,"identity":"5d3601c0-f14d-4cd1-9adc-9e310d0f36f5","order_by":0,"name":"Jinghao Xue","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAqElEQVRIiWNgGAWjYBACxhkgsiKBgQ1IGZCg5QwpWhgkQPraEkhwGPPsHjOJj/PS5PjYGxiKC4hy2JxjaZIzt+UYs/EcYDCeQZSWGcnHpHm3VSS2SSQwGPMQpyWxTfrvnIp6UrQAbWFsyElgI0FLWrJlz7E0wzaegw3E+cVwRo7hjR81yfLy7c3HjIkKMcMGhIVtxsToYJBHYjM/JkrLKBgFo2AUjDgAAG/XLG8BVwXBAAAAAElFTkSuQmCC","orcid":"","institution":"Keio University","correspondingAuthor":true,"prefix":"","firstName":"Jinghao","middleName":"","lastName":"Xue","suffix":""},{"id":554856592,"identity":"d18318bd-1d1f-4ec0-aa6a-1c4e18f727e6","order_by":1,"name":"Tianxiang Yang","email":"","orcid":"","institution":"Keio University","correspondingAuthor":false,"prefix":"","firstName":"Tianxiang","middleName":"","lastName":"Yang","suffix":""},{"id":554856593,"identity":"9da0e7b5-f95e-4262-bb08-20ecb82b95e0","order_by":2,"name":"Hideo Suzuki","email":"","orcid":"","institution":"Keio University","correspondingAuthor":false,"prefix":"","firstName":"Hideo","middleName":"","lastName":"Suzuki","suffix":""}],"badges":[],"createdAt":"2025-11-14 12:53:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8114131/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8114131/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97480801,"identity":"cabc19a2-c19c-4e34-9c22-51596a2fcb41","added_by":"auto","created_at":"2025-12-04 21:39:59","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5310,"visible":true,"origin":"","legend":"","description":"","filename":"85e173a3119a4f49a8fb1e4dd749d83f.json","url":"https://assets-eu.researchsquare.com/files/rs-8114131/v1/cb39a224542c32269e1430e0.json"},{"id":99319002,"identity":"8e7c70c4-7d33-4922-b5f7-bf064e0d10fd","added_by":"auto","created_at":"2025-12-31 16:35:58","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2764090,"visible":true,"origin":"","legend":"","description":"","filename":"ArticleTitle.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8114131/v1_covered_800831c3-cc13-4775-9e11-2c048ebfdb31.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"H-PLE: Algorithmic modeling of recommendation systems based on hierarchical PLE","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":"Multitask Learning, Recommendation Systems, Hierarchical Modeling, Negative Transfer","lastPublishedDoi":"10.21203/rs.3.rs-8114131/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8114131/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"In this study, we propose H-PLE (Hierarchical Progressive Layered Extraction), a novel multi-task \nrecommendation framework designed to alleviate negative transfer and enhance performance under extreme \nabel imbalance. By introducing a hierarchical expert structure combined with improved gating and feature \ninteraction modules, H-PLE effectively balances shared and task-specific knowledge extraction. We evaluate \nthe model on large-scale real-world datasets and compare it against established baselines such as MMoE \nand PLE. The experimental analysis demonstrates that H-PLE consistently achieves stronger ranking performance, \nparticularly in sparse-label tasks, while maintaining stable optimization behavior across multiple runs. \nMoreover, we reveal a practical gap between high ranking metrics and low decision-level metrics under \ndefault thresholding, and provide solutions through probability calibration, validation-based threshold \nselection, and top-K or rate-controlled strategies. These findings not only highlight the methodological \nadvantages of H-PLE but also contribute actionable insights for bridging the gap between offline evaluation \nand online deployment in large-scale recommendation systems.","manuscriptTitle":"H-PLE: Algorithmic modeling of recommendation systems based on hierarchical PLE","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-04 21:39:54","doi":"10.21203/rs.3.rs-8114131/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":"787d34b5-13d8-4030-b20f-73e493c0a35b","owner":[],"postedDate":"December 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-12-30T12:40:58+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-04 21:39:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8114131","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8114131","identity":"rs-8114131","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}