Migrating Traditional Database-Based Systems onto Semantic Layer

Across various domains, the ontologies are learned from relational models of databases for the purpose of integrating data to a semantic layer. Through mappings among ontologies, new semantic applications can visit these integrated data with semantic query language like SPARQL. However, the traditional systems are still running on databases with SQL. In this paper we analyze the translations from basic relational algebra operations to SPARQL queries under an ontology learning approach. The result shows that SPARQL is relational completeness in the scenario. So the database-based application could be migrated from relational database onto semantic layer by replacing SQLs with SPARQL queries.