R2LD: Schema-based Graph Mapping of relational databases to Linked Open Data for multimedia resources data

Abstract

The Web of Data used to share and interchange the diverse data of heterogeneous types on the Web has been actively established. Ontology-based Linked Open Data (LOD) that allows computers to understand and process data semantics has emerged to extend the current Web of Documents. LOD is important for the modeling of multimedia resources since it provides an efficient way for unstructured data resources. This paper proposes a noble and practical schema-based mapping method to populate Linked Open Data sets from relational databases for multimedia resources. The proposed schema-based mapping R2LD realizes the seamless mapping for RDB-to-RDF by taking advantage of the compatible conceptual schema. The schema-based mapping can resolve the complicated mapping issues in RDB-to-RDF, such as primary and foreign key relationships. The mapping description is straightforward and flexible. It can define mapping information in the form of attribute-value pairs. Especially, the proposed mapping method is suitable for the unstructured multimedia resources. The schema-based mapping R2LD provides an efficient way to implement SPARQL endpoint into RDB and preserve the performance of SQL, which is vital to the dissemination of LOD.