CC BY 4.0Rhee, Hea Lim2026-09-232026-09-232026-09-23https://2026.ifla.orghttps://repository.ifla.org/handle/20.500.14598/7404When 23 government-funded research institutes produce data under different metadata practices, how does a national platform become a trustworthy participant in global knowledge exchange? This poster presents DataON’s redesigned three-layer metadata standard, developed for Korea’s national research data platform operated by KISTI, serving 2.95 million records, and holder of Korea’s first CoreTrustSeal certification. Rather than a single rigid schema, the architecture combines a national Canonical Core, a FAIR and quality-maturity layer, and governed domain profiles. Layer 1 defines a canonical model built on DataCite 4.7 with full Dublin Core mapping, unconditional and conditional mandatory rules, and a unified persistent-identifier policy connecting datasets, researchers, institutions, and projects. Layer 2 organizes elements into Bronze–Silver–Gold–Platinum maturity tiers, extending toward AI-ready, machine-actionable metadata. Layer 3 operates community-developed domain profiles—geoscience, life sciences, social sciences—through formal approval, core-mapping, and validation rules, aligned with DCAT, OAI-PMH, and PROV-O. DataON’s experience shows that metadata standardization is not merely schema alignment but a socio-technical process—one that lets a national platform evolve into a living, verifiable standard and a genuine bridge to global knowledge infrastructure.enghttps://creativecommons.org/licenses/by/4.0/MetadataResearch dataFAIR dataInteroperabilityData mappingsResearch data repositoriesPostersHea Lim Rhee