Abstrakt
RDF is a universal data model for publishing structured data on the Web. On the other hand, many structured data is stored in relational database systems.
To support publishing data in the RDF model, it is essential to close the gap between the relational and RDF worlds. A virtual SPARQL endpoint over relational data is a promising approach to achieve that.
To build a virtual SPARQL endpoint, we need to know how to translate SPARQL queries to corresponding SQL queries. There exist several approaches to such transformation.
Most of them are focused on the processing of user-defined mapping. The user-defined mapping gives an user the ability to define a mapping of a stored relation data to almost any RDF representation.
In this paper we focus on one of the core problems of the transformation: how to represent variables from a given SPARQL query in the corresponding SQL query. We survey variable representations from existing approaches; how the selected representation affects the soundness and performance of the whole transformation approach.