A statistical approach for XML query size estimation

The increasing number of XML repositories has intensified research activities in the optimization of XML queries The success of any optimization approach hinges on an accurate query size estimation This paper presents a statistical method for estimating the result size of XML queries Our estimation system extracts two summarized information, namely, node ratio and node factor, from every distinct parent-child path in the XML files Experiment results indicate that our approach requires small memory footprint, and yet proves to be sufficient in estimating the result size of queries under the data-independent assumption.