Assessing the Viability of an OBDA Approach to Integrating Databases into a Virtual Knowledge Graph for Research Management at ITC

Institutional research management depends on data that is often distributed across independent operational systems. At the Institute of Technology of Cambodia (ITC), relevant information about students, employees, departments, academic records, researchers, projects, profiles, and publications is spread across several databases. This paper reports an early-stage, applied assessment of an ontology-based data access approach for integrating these sources into a virtual knowledge graph. The implementation uses Trino to federate PostgreSQL databases and Ontop to expose RDF views through a SPARQL endpoint, while preserving the existing source systems. The ontology models core institutional entities and relationships, including cross-system links between employees and researcher identities. Initial mapping demonstrates integration of student academic records and research management. Preliminary SPARQL tests show that the approach can answer integrated questions over academic and research data without copying operational data into a new repository. The work suggests that OBDA is a viable and cost-conscious foundation for institutional knowledge graph development at ITC, while also revealing challenges in identity alignment, schema harmonization, endpoint performance, and long-term governance.