Efficient processing of XML twig queries with OR-predicates

An XML twig query, represented as a labeled tree, is essentially a complex selection predicate on both structure and content of an XML document. Twig query matching has been identified as a core operation in querying tree-structured XML data. A number of algorithms have been proposed recently to process a twig query holistically. Those algorithms, however, only deal with twig queries without OR-predicates. A straightforward approach that first decomposes a twig query with OR-predicates into multiple twig queries without OR-predicates and then combines their results is obviously not optimal in most cases. In this paper, we study novel holistic-processing algorithms for twig queries with OR-predicates without decomposition. In particular, we present a merge-based algorithm for sorted XML data and an index-based algorithm for indexed XML data. We show that holistic processing is much more efficient than the decomposition approach. Furthermore, we show that using indexes can significantly improve the performance for matching twig queries with OR-predicates, especially when the queries have large inputs but relatively small outputs.