XML document summarization: using XQuery for synopsis creation

This work presents a methodology to support approximate queries over massive and heterogeneous XML data sets, based on concise data statistics such as histograms or other statistical techniques. The basic idea for approximate answers is to store precomputed summaries of the XML data, also called synopses, and to query them instead of the original database, thus saving time and computational costs. In particular, We concentrate on a set of XQuery transformation rules for the construction of the synopses collection and for querying the synopsis.