Scientific workflows increasingly produce structured JavaScript Object Notation (JSON) data. Although this data is easy to exchange, it is still hard to interpret consistently across systems at the semantic level. JSON Schema helps with structural validation, but it does not provide native Linked Data semantics. This thesis addresses that gap by extending MetaConfigurator, a schema-driven web application for structured data authoring, with Semantic Web capabilities.

The main contribution is an integrated Resource Description Framework (RDF) Authoring Panel that supports end-to-end semantic data workflows in a single user interface. Researchers can start from existing hierarchical or tabular research data, use the AI-assisted RDF Mapping Language (RML) generation to transform it into RDF, and then continue in the same environment to refine triples, run SPARQL queries, visualize knowledge graph, and export to common RDF serializations.

Overall, this work presents a practical end-to-end approach for transforming structured data into semantically enriched representations. A case study is conducted using laboratory data on Metal-Organic Framework (MOF) synthesis in chemistry. In this study, protocol data is converted from JSON into ontology-based JSON-LD using RML, where synthesis runs, reagents, procedure steps, and quantities are explicitly linked through well-defined classes and relationships. Within the same interface, users can progressively refine the generated triples, benefit from ontology-aware IRI suggestions, create and adjust SPARQL queries with AI assistance, and visually explore and edit the resulting knowledge graph. This integrated workflow preserves the original experimental context while making the data easier to query, compare across runs, integrate with external knowledge systems, and reuse in future research.