Data quality support to on-the-fly data integration using adaptive query processing

In dynamic, on-the-fly relational data integration settings, such as data mashups, there is a need to reconcile values heterogeneity across sources, in order to ensure consistency and completeness of the integrated data. In this scenario, the use of exact joins to match records across sources may lead to incomplete integration, while similarity joins, often advocated as a solution to this problem, is computationally expensive. In this paper we explore the use of adaptive query processing (AQP) techniques in order to combine exact (fast) and approximate (accurate) joins when performing dynamic integration. The adaptive algorithm uses an an priori expectation of the join result size combined with the monitoring of join progress to statistically determine, at various points during query execution, which join operator should be used. Depending on its configuration, the algorithm can achieve various tradeoffs between completeness of the join result, and query execution time. Our experimental results show that sensible savings in join execution time can be achieved in practice, at the expense of a modest reduction in result completeness.