Dynamic Query Routing with Aleatoric and Epistemic Uncertainty Handling for Virtual Assistants: A Hybrid Approach in Retrieval-Augmented Generation

Dynamic Query Routing with Aleatoric and Epistemic Uncertainty Handling for Virtual Assistants: A Hybrid Approach in Retrieval-Augmented Generation | 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 Query Routing with Aleatoric and Epistemic Uncertainty Handling for Virtual Assistants: A Hybrid Approach in Retrieval-Augmented Generation Ayush Giri, Rabindra Adhikari, Abhinav Giri, Prof. Saumendra Mohanty This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7201693/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 Virtual assistants need effective query processing in orderto provide precise responses in uncertain, knowledge-rich scenarios.This work introduces a hybrid system for dynamic query routing anduncertainty-sensitive response generation in a retrieval-augmented generation approach. The system combines query embedding, classification,and uncertainty management in routing queries over heterogeneousknowledge sources, with subsequent response generation. The systemcounters aleatoric uncertainty through the analysis of semantics inthe query, with epistemic uncertainty managed through confidencecalibration, supporting robust performance. The evaluations yield91 percent routing accuracy in routing, an Expected Calibration Error(ECE) of 0.06, a 95 percent Uncertainty-Handled Query Success Rate(UH-QSR), and an average response time of 0.55 s, achieving 15 percentimprovement in accuracy as well as 50 percent speedup over thebaseline. The system’s efficiency and resilience are supported throughnovel quantities of a Routing Efficiency Index (REI) of 18.2 and Quality Efficiency Ratio (QER) of 1.60. The presented approach contributes tovirtual assistant systems with scalable solutions for conversational AIapplications. Query Processing Aleatoric Uncertainty Epistemic Uncertainty Virtual Assistants Retrieval-Augmented Generation Conversational AI Dynamic Routing Uncertainty Handling Routing Efficiency Index Quality-Efficiency Ratio 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-7201693","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":498253482,"identity":"cd927370-7a5c-4d29-aff0-1f1639c46664","order_by":0,"name":"Ayush Giri","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIiWNgGAWjYJACZgYGCQY29gYgo4AkLTwHgAwD4rUAgUQCkVr4Z+QYfy74ZZHPJ/n24OcCAwZ5frEDbI/xOVDiRo6Z9Mw+Ccs26bxk6RkGDIYzZyewG8/AZw1QCzNvj4QBmzSQwWPAkGBwO4FNmgePDvkbQIeBtUieIVKLwY0cA2meH0AtEjxEajE886xMmrcBqIUnx1iax0AC6JfENml8fpE7nrz5M8+fOgP59jOGn3kqbOT5pZOPSeONUoEEBgbGNjhXAogZG5jx6WDgPwAk/qAJ4tcyCkbBKBgFIw0AAP6CO4b5BYgjAAAAAElFTkSuQmCC","orcid":"","institution":"Nepal Clearing House Ltd","correspondingAuthor":true,"prefix":"","firstName":"Ayush","middleName":"","lastName":"Giri","suffix":""},{"id":498253485,"identity":"8d814588-b80e-4b27-8554-7bd2adfaba2a","order_by":1,"name":"Rabindra Adhikari","email":"","orcid":"","institution":"Nepal Clearing House Ltd","correspondingAuthor":false,"prefix":"","firstName":"Rabindra","middleName":"","lastName":"Adhikari","suffix":""},{"id":498253486,"identity":"41fff412-e57c-4ce9-a59e-de9ca717029e","order_by":2,"name":"Abhinav Giri","email":"","orcid":"","institution":"Vellore Institute of Technology University","correspondingAuthor":false,"prefix":"","firstName":"Abhinav","middleName":"","lastName":"Giri","suffix":""},{"id":498253488,"identity":"b2494dc9-9eb1-4228-a7f8-4a70054a9750","order_by":3,"name":"Prof. Saumendra Mohanty","email":"","orcid":"","institution":"Bennett University","correspondingAuthor":false,"prefix":"","firstName":"Prof.","middleName":"Saumendra","lastName":"Mohanty","suffix":""}],"badges":[],"createdAt":"2025-07-24 05:38:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7201693/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7201693/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93352003,"identity":"02eea3ff-2bd3-4a74-abbd-d94244cddf56","added_by":"auto","created_at":"2025-10-12 20:31:36","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1280246,"visible":true,"origin":"","legend":"","description":"","filename":"Uncertanity.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7201693/v1_covered_16fb98fc-6fa5-4e9a-923e-bdb99a6f6580.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Dynamic Query Routing with Aleatoric and Epistemic Uncertainty Handling for Virtual Assistants: A Hybrid Approach in Retrieval-Augmented Generation","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":"Query Processing, Aleatoric Uncertainty, Epistemic Uncertainty, Virtual Assistants, Retrieval-Augmented Generation, Conversational AI, Dynamic Routing, Uncertainty Handling, Routing Efficiency Index, Quality-Efficiency Ratio","lastPublishedDoi":"10.21203/rs.3.rs-7201693/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7201693/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Virtual assistants need effective query processing in orderto provide precise responses in uncertain, knowledge-rich scenarios.This work introduces a hybrid system for dynamic query routing anduncertainty-sensitive response generation in a retrieval-augmented generation approach. The system combines query embedding, classification,and uncertainty management in routing queries over heterogeneousknowledge sources, with subsequent response generation. The systemcounters aleatoric uncertainty through the analysis of semantics inthe query, with epistemic uncertainty managed through confidencecalibration, supporting robust performance. The evaluations yield91 percent routing accuracy in routing, an Expected Calibration Error(ECE) of 0.06, a 95 percent Uncertainty-Handled Query Success Rate(UH-QSR), and an average response time of 0.55 s, achieving 15 percentimprovement in accuracy as well as 50 percent speedup over thebaseline. The system’s efficiency and resilience are supported throughnovel quantities of a Routing Efficiency Index (REI) of 18.2 and Quality Efficiency Ratio (QER) of 1.60. The presented approach contributes tovirtual assistant systems with scalable solutions for conversational AIapplications.","manuscriptTitle":"Dynamic Query Routing with Aleatoric and Epistemic Uncertainty Handling for Virtual Assistants: A Hybrid Approach in Retrieval-Augmented Generation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-11 04:34:15","doi":"10.21203/rs.3.rs-7201693/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":"214f4527-fc34-454c-a51e-45ce8ff88f98","owner":[],"postedDate":"August 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-12T20:23:26+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-11 04:34:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7201693","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7201693","identity":"rs-7201693","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}