Automatic Generation of Query Order Execution Plan for Hybrid Mediator with Medical Sources

In the context of heterogeneous integration systems the choice of data sources that respond to user requests is usually a very difficult task. It is precisely because the data sources can often be replaced by overlapping and some data transfer are unpredictable and not useful. Consuming entirely or partially the data produced by set of sources before accessing another one, increase the cost of total response time. So, the query engine must take account the order of the relevant sources to retrieve the data. The selection of the sources and the order plan of query execution is therefore crucial for obtaining high-quality and good performance. In this paper we present a new architecture of hybrid mediation with medical sources. Whose core functionality is based on combination between a relevant components of data warehouse and those of virtual mediator. Thereafter, we propose a new and innovate method for query order execution plan in hybrid data integration system. The solution designed to support adaptivity at its core using a two-pronged approach. Similarity of the data sources and partial semantic relationship between sources based on calculate inaccurate estimates. We demonstrate that the solution proposed extends the related work in adaptive execution. Therefore, we present experimental evidence that demonstrate our algorithms provide the interesting result for a data integration system.