Fragmenting XML Documents via Structural Constraints

XML query processors suffer from main-memory limitations that prevent them from processing large XML documents. While content-based predicates can be used to project down parts of the documents, it may still be needed to resize the obtained projections according to structural constraints. In this paper, we consider size, tree-width and tree-depth constraints to enable a structuredriven fragmentation of XML documents. Although a set of heuristics performing this kind of fragmentation can be easily devised, a key problem is determining the values of structural constraints input to the above heuristics, given that the search space is prohibitive at large. To alleviate the problem, we introduce special-purpose structure histograms that report the constraint values for the fragments of a given document. We then present a prediction algorithm that probes those histograms to output the expected number of fragments, when fixed input values of the constraints are used. Furthermore, we study how to relax the fixed constraints by means of classical distributions. An experimental evaluation of our study shows the effectiveness of our fragmentation methodology on some representative XML datasets.