Neuro-Symbolic AI for Hybrid Life Cycle Assessment under Missing Not At Random Data: Inventory Completion from Electronic Tax Records

Neuro-Symbolic AI for Hybrid Life Cycle Assessment under Missing Not At Random Data: Inventory Completion from Electronic Tax Records | 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 Neuro-Symbolic AI for Hybrid Life Cycle Assessment under Missing Not At Random Data: Inventory Completion from Electronic Tax Records Sakayong Pattanavekin, Sanong Ekgasit This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9052200/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 Hybrid life cycle assessment (LCA) from procurement ledgers fails first as an inventory-completion problem. Standardized electronic tax records are auditable, but they primarily record expenditure, supplier identi-fiers, dates, and commodity tags rather than the mass and energy flows required for process-based inventory construction. Because invoices without physical units cluster among suppliers with weaker reporting in- frastructures, the resulting gaps are Missing Not At Random (MNAR). Static environmentally extended input-output (EEIO) factors recover completeness by suppressing within-category variance. Unconstrained tabular machine learning improves fit, but ignores supply-chain dependence and offers no auditable uncer-tainty contract. This paper presents a dual-pathway neuro-symbolic AI framework for hybrid LCA on e-TAX records. Path A performs deterministic activity-based calculation when physical units are observed. Path B estimates spend-normalized cradle-to-gate intensity for spend-only records using semantic priors, graph-attention message passing over supplier topology, organizational-intensity anchors, a deployment-scale macro-boundary safeguard, and split-conformal calibration. Empirical evaluation uses a product-level proxy design built from 132 third-party verified cement and concrete footprints in the Thailand Greenhouse Gas Management Organization (TGO) registry, with 42 observations for fitting and 90 for blinded testing. On the test set, the proposed framework reduces RMSE from 0.185 to 0.042 kg CO2e/THB relative to a static EEIO baseline and attains 95.2% empirical coverage at a 90% nominal target. The results show that digital public infrastructure can anchor hybrid LCA under MNAR when inference is constrained, graph-aware, and calibrated. hybrid life cycle assessment Missing Not At Random neuro-symbolic AI digital public infrastructure e-TAX conformal prediction 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-9052200","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":608834717,"identity":"3693f572-a0c5-4f4a-b9b8-a141199a7992","order_by":0,"name":"Sakayong Pattanavekin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABI0lEQVRIie3PMUvDQBTA8RcCyXIm6+uUr3AQiBlK81XuCMSluBSKINRA4bro7qRfwS7B8eSgWQ67BroogpNLKUjdTBsVhys4OtwfLpBLfrkXAJvtn+cCgz4E3Q3uLvTgu5L9kAK8L4J/IwDqm8BBEpVHD5v1dnAaXueKPt8vM8+fvj45w3QC/rRCA6EyyFGyfIRNkTOuV1yQxTF1qnYwshgbCRDaEpeXWseSixXzkCW4JzhMTCQqSbyV7ILf6uWmJY+ZhyfvHYnejAQkSdpTFL+rL13GhXTE7uPdKcRIqAqKVBc1n9ciplzk7b/oMfIKe4IUo9Q02OxKNWf9c36j3JfehxhkoT+rcF1NwtBX88Y0mEtMu6xdnunBPiOx2Ww2268+Acg9XVOwhNuMAAAAAElFTkSuQmCC","orcid":"","institution":"Chulalongkorn University","correspondingAuthor":true,"prefix":"","firstName":"Sakayong","middleName":"","lastName":"Pattanavekin","suffix":""},{"id":608834718,"identity":"229d0f15-b7d2-448c-b319-2107d7fd316e","order_by":1,"name":"Sanong Ekgasit","email":"","orcid":"","institution":"Chulalongkorn University","correspondingAuthor":false,"prefix":"","firstName":"Sanong","middleName":"","lastName":"Ekgasit","suffix":""}],"badges":[],"createdAt":"2026-03-06 15:38:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9052200/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9052200/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106333710,"identity":"51d5de6f-4a24-4f1e-b034-b541cffaec4a","added_by":"auto","created_at":"2026-04-07 14:28:14","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":348372,"visible":true,"origin":"","legend":"","description":"","filename":"VEKINJIEEdition.tex.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9052200/v1_covered_4c1c9fc9-2109-4298-bdfb-cd721479cfaf.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Neuro-Symbolic AI for Hybrid Life Cycle Assessment under Missing Not At Random Data: Inventory Completion from Electronic Tax Records","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":"hybrid life cycle assessment, Missing Not At Random, neuro-symbolic AI, digital public infrastructure, e-TAX, conformal prediction","lastPublishedDoi":"10.21203/rs.3.rs-9052200/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9052200/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHybrid life cycle assessment (LCA) from procurement ledgers fails first as an inventory-completion problem.\u0026nbsp;Standardized electronic tax records are auditable, but they primarily record expenditure, supplier identi-fiers, dates, and commodity tags rather than the mass and energy flows required for process-based inventory\u0026nbsp;construction. Because invoices without physical units cluster among suppliers with weaker reporting in-\u0026nbsp;frastructures, the resulting gaps are Missing Not At Random (MNAR). Static environmentally extended input-output (EEIO) factors recover completeness by suppressing within-category variance. Unconstrained\u0026nbsp;tabular machine learning improves fit, but ignores supply-chain dependence and offers no auditable uncer-tainty contract.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis paper presents a dual-pathway neuro-symbolic AI framework for hybrid LCA on e-TAX records.\u0026nbsp;Path A performs deterministic activity-based calculation when physical units are observed. Path B estimates\u0026nbsp;spend-normalized cradle-to-gate intensity for spend-only records using semantic priors, graph-attention message passing over supplier topology, organizational-intensity anchors, a deployment-scale macro-boundary\u0026nbsp;safeguard, and split-conformal calibration. Empirical evaluation uses a product-level proxy design built from\u0026nbsp;132 third-party verified cement and concrete footprints in the Thailand Greenhouse Gas Management Organization (TGO) registry, with 42 observations for fitting and 90 for blinded testing. On the test set, the\u0026nbsp;proposed framework reduces RMSE from 0.185 to 0.042 kg CO2e/THB relative to a static EEIO baseline and\u0026nbsp;attains 95.2% empirical coverage at a 90% nominal target. The results show that digital public infrastructure\u0026nbsp;can anchor hybrid LCA under MNAR when inference is constrained, graph-aware, and calibrated.\u003c/p\u003e","manuscriptTitle":"Neuro-Symbolic AI for Hybrid Life Cycle Assessment under Missing Not At Random Data: Inventory Completion from Electronic Tax Records","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-20 06:50:51","doi":"10.21203/rs.3.rs-9052200/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":"3440ab25-16ec-4e09-9966-3b2f5d549ad3","owner":[],"postedDate":"March 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-14T13:24:50+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-20 06:50:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9052200","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9052200","identity":"rs-9052200","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}