The contemporary ecosystem of digital repositories is characterized by significant heterogeneity among objects or collections of objects, software, and services, as well as a growing expectation that knowledge will be more openly available, reusable, and machine-actionable. A theoretical framework is developed in the present paper to address the issue of metadata harmonization and semantic interoperability in digital repositories. Its utility is demonstrated through a simulated academic dataset of 60 repositories representing universities, libraries, archives, museums, and research data platforms. This study explores the relationships of metadata completeness, controlled vocabularies, persistent identifiers, availability of OAI-PMH/APIs, ontology or linked-data support and repository maturity with resource discovery, data exchange efficiency and long term access. The framework was evaluated through the use of descriptive statistics, cross-tabulation, correlation analysis, regression analysis, ANOVA and reliability testing. According to simulated results, discovery effectiveness is highly associated with metadata completeness, while persistent identifier, API availability, and linked-data integration have substantial contributions to long-term accessibility. Repositories employing mixed schema or domain responsive metadata practices tend to yield stronger interoperability performance than those using isolated descriptive fields without machine-readable semantic structures. The findings of our analysis indicate that metadata harmonization is by no means a simple schema conversion task. It entails governance, semantic mapping, quality control, preservation metadata, authority control and ongoing technical maintenance. A framework is offered to repository managers, metadata librarians, digital archivists, and information policy makers which can support better discoverability, reduce (meta)data fragmentation, and foster durable access to digital resources.
Kumar et al. (Tue,) studied this question.