Augmenting Large Language Models for Enhanced Interaction with Government Data Repositories

Augmenting Large Language Models for Enhanced Interaction with Government Data Repositories | 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 Augmenting Large Language Models for Enhanced Interaction with Government Data Repositories Paul Trust, Kizito Omala, Rosane Minghim This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3897706/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 the modern digital landscape, government agencies globally are shifting services online to enhance transparency and public engagement. However, the vast digital content can be daunting for citizens seeking information. Addressing this, our research evaluates the efficacy of Large Language Models (LLMs), like ChatGPT, in the public sector, highlighting their potential in extracting relevant insights and optimizing information navigation. Our approach integrates non-parametric data from various sources focusing on information posted on three irish websites; the government publications, health services, and the Citizens Information websites, using retrieval-augmented models. Empirical evaluations show that the llama2 model, with $13$ billion parameters, achieves up to 90% for government publication releases and up to 96.12% for health information enhancement when complemented with retrieval augmentation, with other models also showing substantial improvements. These results emphasize the transformative potential of retrieval-augmented frameworks in keeping LLMs updated with the evolving public information domain. Computational Mathematics Applied Statistics Augmented Learning Language Models Public Sector Full Text Additional Declarations The authors declare no competing interests. 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-3897706","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":269276116,"identity":"255a55b5-ff80-4018-b153-bb053ad97b6b","order_by":0,"name":"Paul Trust","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIie3PoW4CMRjA8V5Iivl2ugRSXuFbyC6Ie5ieqUJMLYiJwxyGB5jbK0xNNzThTB+AC4g7g0IgScYIPZYMwSjgJvoXTdr0l34lxOf7h9HmaFRuhjHQZmq32P85Fg4SgtaPb0byEFRN2HXCmZTth2za4+x4jd0wGBkgA6qSrLV++tg+M4K5Lkk5dBGD1iyTrD2IiokdDI1EIoyDBBMUyFZHMoeazIGQJHOQBqASqO1gJip2v2TvIJQmqRK6RxlEi9MrqYNAQwepkpyCfFl07Lda9i9KzC6T7ns1/vrex9Ad689ivYt5mOuq2r5eJmdBvag7gM/n8/n+6AANdk6VkZ9MVQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-0177-1096","institution":"University College Cork, Cork, Ireland","correspondingAuthor":true,"prefix":"","firstName":"Paul","middleName":"","lastName":"Trust","suffix":""},{"id":269276118,"identity":"ad30fc28-8334-4e03-9256-63407629077d","order_by":1,"name":"Kizito Omala","email":"","orcid":"","institution":"Makerere University, Kampala, Uganda","correspondingAuthor":false,"prefix":"","firstName":"Kizito","middleName":"","lastName":"Omala","suffix":""},{"id":269276117,"identity":"fd57fc7e-8104-4ec3-89f7-4c8491888c5e","order_by":2,"name":"Rosane Minghim","email":"","orcid":"","institution":"University College Cork, Cork, Ireland","correspondingAuthor":false,"prefix":"","firstName":"Rosane","middleName":"","lastName":"Minghim","suffix":""}],"badges":[],"createdAt":"2024-01-25 15:57:23","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-3897706/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3897706/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":52776034,"identity":"2b2bfc87-0c92-4b14-bbb3-6861443618b5","added_by":"auto","created_at":"2024-03-15 15:50:07","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":641235,"visible":true,"origin":"","legend":"","description":"","filename":"interactiveconversationLLMforpublicsectorReadyforReview8.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3897706/v1_covered_d6125ccc-e0a3-4073-bbdd-5123c3cf3932.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eAugmenting Large Language Models for Enhanced Interaction with Government Data Repositories\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University College Cork","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":"Augmented Learning, Language Models, Public Sector","lastPublishedDoi":"10.21203/rs.3.rs-3897706/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3897706/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the modern digital landscape, government agencies globally are shifting services online to enhance transparency and public engagement. However, the vast digital content can be daunting for citizens seeking information. Addressing this, our research evaluates the efficacy of Large Language Models (LLMs), like ChatGPT, in the public sector, highlighting their potential in extracting relevant insights and optimizing information navigation. Our approach integrates non-parametric data from various sources focusing on information posted on three irish websites; the government publications, health services, and the Citizens Information websites, using retrieval-augmented models. Empirical evaluations show that the llama2 model, with $13$ billion parameters, achieves up to 90% for government publication releases and up to 96.12% for health information enhancement when complemented with retrieval augmentation, with other models also showing substantial improvements. These results emphasize the transformative potential of retrieval-augmented frameworks in keeping LLMs updated with the evolving public information domain.\u003c/p\u003e","manuscriptTitle":"Augmenting Large Language Models for Enhanced Interaction with Government Data Repositories","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-26 16:31:06","doi":"10.21203/rs.3.rs-3897706/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":"1a80097f-dbf0-48f3-92a2-de3485d1e7b5","owner":[],"postedDate":"January 26th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":28374927,"name":"Computational Mathematics"},{"id":28374928,"name":"Applied Statistics"}],"tags":[],"updatedAt":"2024-01-26T16:31:06+00:00","versionOfRecord":[],"versionCreatedAt":"2024-01-26 16:31:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3897706","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3897706","identity":"rs-3897706","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}