You are here


ELITE is a federated query engine that facilitates a complete and transparent querying of distributed autonomous data sources, for example relational databases or triple stores. ELITE uses SPARQL queries that refer to an ontology which is univocally mapped onto the schemas of data sources.

At runtime the queries are translated into the query language (usually SQL) comprehensible for the data sources. This approach simplifies the query design as long as the end-user does not have to know how the physical data layer is organized, formatted, and structured.

The query rewriting and execution is controlled by a semantic index look up procedure that helps to identify data available at each single federated data source. The query answers provided by these sources are fused and delivered to the application which has formulated the original query.

In other words ELITE can be seen as a virtual semantic data base that provides a homogeneous querying procedure to data physically stored at different locations and aligned at different schema. For the end user ELITE hides the complexity of the querying process. 

From the technical point of view ELITE combines several different technologies. It makes use of Ontology based data access (OBDA) techniques that facilitate query rewriting.  On the one hand SPARQL to SPARQL, by taking into account the preselected entailment regimes and by these means to guarantee completeness and soundness of query answers. On the other hand SPARQL to SQL to enable accessing of relational data sources. Beside that, ELITE supports highly efficient query processing by means of an improved implementation of R-Tree-based indexing.

Apart of its direct use as a central SPARQL end-point service for virtual data bases, ELITE can be applied for evaluation of consistency in distributed data repositories.