Overcoming Limitations of Sampling for Aggregation Queries

We study the problem of approximately answering aggregation queries using sampling. We observe that uniform sampling performs poorly when the distribution of the aggregated attribute is skewed. To address this issue, we introduce a technique called outlier-indexing. Uniform sampling is also ineffective for queries with low selectivity. We rely on weighted sampling based on workload information to overcome this shortcoming. We demonstrate that a combination of outlier-indexing with weighted sampling can be used to answer aggregation queries with significantly reduced approximation error compared to either uniform sampling or weighted sampling alone. We discuss the implementation of these techniques on Microsoft’s SQL Server, and present experimental results that demonstrate the merits of our techniques.